Design Of Water Quality Monitoring System Based On Small USV

With the rapid development of China’s economy and society, exploration and usage of water resources is increasingly causing concern and attentions. However, the existing water quality monitoring systems which require a lot of manpower and material to support, are always too big and expensive for wide range of applications. In response to this social demand, by summing up experience and technology in related fields, we designed water quality monitoring system, which is suitable for coastal waters and lakes. The system consists of two parts, the USV on water is responsible for data acquisition, and the base station onshore is mainly for data processing. The system is miniaturized, intelligent, and modular designed, which is easy to use and to maintain.With an Android operating system running on it, the USV control center can provide good visual operating experience in the development and design process. The single beam sonar, GPS, and AHRS on board ensures the USV to navigate safely. The modular designed sensor group provide a variety of sensor interface, and we can choose suitable sensors according to actual need.In this paper, we focused on the key technologies used in the system, and conducted a verification experiment in a lake. To ensure the safety of navigation, USV requires real-time obstacle avoidance capability. On the basis of in-depth analysis of sonar control commands, we designed the obstacle avoidance algorithm whose detection ranges vary according to the distance between the sonar and the obstacle, while we divide the terrain contour into three categories, and describe it with mathematical formulas. For each type of terrain, we use fuzzy control technology to make the obstacle avoidance algorithm more intelligent.In order to analyze the underwater environment precisely, we take a number of steps to get the underwater terrain. Firstly, we obtain the two-dimensional polar coordinate slices with a single beam sonar, and convert it to terrain slices under Cartesian coordinate system through coordinate transformation. Secondly, we combine the two-dimensional information to three-dimensional point cloud, and process it with the PCL point cloud filtering library to make it smooth and regular. Finally, with the help of blender we use the Cover---Contraction method to obtain the underwater terrain model with high accuracy.We take the water temperatures a representative parameter to conduct the experiment in the lake, and analysis the coordinate information and temperature data. The experiment achieved the desired goal and verified the feasibility and practicality of the system.