Research on Multi-sensor Fusion Perception Method of Vehicle-infrastructure Collaboration for Smart Automobiles

Abstract

Traffic congestion should be solved, it reduce traffic safety potential risks, and improve people's travel efficiency. The paper based on intelligent car's own operation characteristics (Smart car perception is generally only 300 ~ 500 meters, there is a perceived problem of super-visual distance and vision blind zone), In order to effectively improve the safety operation level of smart cars, and combined with intelligent non-smart cars on the road may encounter cars a smart car trip perception highway outside the scope of frequently asked questions, such as over-site sensations, the whole scene and area's perception, the blind area, the emergency corner, tunnel, bridge, other highways travel common scenes and so on. This paper is based on new infrastructure transformations or newly build's research and practical results such as road traffic intelligence infrastructure, it deployed the current road traffic to intelligently infrastructure, especially the technical difficulties existing in the process of common perceptual equipment (smart cameras, radar) are synonymous with multiple sensor information fusion perceptions, it proposed a multi-sensor fusion algorithm based on error variance, and designed a multi-object multi-sensor data processing system architecture. This paper also proposes traffic operation scheduling architecture based on the game theory of car road synergies on the basis of multi-sensor data fusion. Finally, these architectures were analyzed using computer simulation techniques. The results show that the traffic operation schedule for multi-sensor fusion algorithm based on error variance and game theory based on the study proposed this study can be more obvious. The safety and efficiency of the road traffic environment of smart vehicles. Optimization, smart vehicles equipped with smart vehicles are also more than 25% higher than traditional common vehicles in terms of vehicle safety. All in all, this study proposed to synergistic multi-sensor convergence method for smart cars, that based on smart car a smart car perception, compared to non-smart roads after intelligent infrastructure construction and transformation of road traffic intelligent transportation systems, Higher efficiency and more intelligent can better solve the common problems in road traffic environment, providing people with safer, efficient and high-quality traffic travel services.