Abstract
Patrol unmanned aerial vehicles (UAVs) in coal mines have high requirements for environmental perception. Because there are no GPS signals in a mine, it is necessary to use simultaneous localization and mapping (SLAM) to realize environmental perception for UAVs. Combined with complex coal mine environments, an integrated navigation algorithm for unmanned helicopter inertial measurement units (IMUs), light detection and ranging (LiDAR) systems, and depth cameras based on probabilistic membrane computing-based SLAM (PMC-SLAM) is proposed. First, based on an analysis of the working principle of each sensor, the mathematical models for the corresponding sensors are given. Second, an algorithm is designed for the membrane, and a probabilistic membrane system is constructed. The probabilistic SLAM map is constructed by sparse filtering. The experimental results show that PMC can further improve the accuracy of map construction. While adapting to the trend of intelligent precision mining in coal mines, this approach provides theoretical support and application practice for coal mine disaster prevention and control.
Highlights
unmanned aerial vehicles (UAVs) application scenarios have high requirements for navigation performance, especially in complex environments, and a single navigation mode has difficulty meeting these requirements
In reference [3], UAV simultaneous localization and mapping (SLAM) is realized by neural cell population coding, but the simulation can only be completed under specific constraints
A navigation algorithm for UAVs, inertial measurement units (IMUs), light detection and ranging (LiDAR) systems, and depth cameras based on probabilistic membrane computing-based SLAM (PMC-SLAM) is proposed, and the main contributions are as follows: (1) Based on the sensor model, a corresponding mathematical model is given
Summary
UAV application scenarios have high requirements for navigation performance, especially in complex environments, and a single navigation mode has difficulty meeting these requirements. Reference [1] improves the accuracy of data association by using the complementary advantages of point and line features This method is only tested in a simulation environment and lacks training on actual scenes. In reference [9], a visual pose estimation method for underground UAVs based on a depth neural network is proposed. This method can significantly improve the positioning accuracy of roadways in complex environments, but there are some problems, such as large computing resource demands and relatively long computing times. Based on the membrane computing model, an effective membrane algorithm has been designed to solve optimization control problems in various fields of application. (1) Based on the sensor model, a corresponding mathematical model is given (2) A probabilistic membrane system-based calculation model is further constructed (3) An algorithm is designed to realize the map construction process in the membrane
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
