基于 GPS 的自学习导航游弋式水质监测系统设计.

Aquaculture involves cultivating freshwater and saltwater populations under controlled conditions, in which the high water quality plays an important role for the harvest of aquatic organisms. This paper proposes a water quality monitoring system to achieve that goal. While current water quality monitoring devices share drawbacks of small measuring range, poor mobility and high cost, the distinguished contribution of water monitoring is a self-learning navigation component, which can address the previously mentioned challenges in other systems. Our system contains a front-end water monitoring subsystem, as well as a back-end server to store and analyze the monitored data. We developed three main modules in the front-end monitoring subsystem: a water quality collection module, a vessel movement control module, and a GPS navigation module. The water quality collection module contains a PT100 temperature sensor, a fluorescence dissolved oxygen sensor, and an industrial pH meter. Those sensors are used to collect parameters related to water quality including water temperature, dissolved oxygen, and the pH value. The vessel motion control is remotely managed by a CC2530 chip, which periodically sends commands to the motion coordinator in the ship. All data from the monitoring subsystem, including the water quality parameters, vessel movement control commands, and the GPS locations, are sent to the GPRS layer, which acts as a bridge to connect the monitoring subsystem and the server. Once the server received data, it parses them and calculates the water temperature, the dissolved oxygen and PH values. Meanwhile, the server extracts the location information and computes the distance and the direction angle to the target position. We have designed a database to store the collected data in the server, and also developed an Android application so that individual users can access the data at all time and places. The user can even set measurement target and control the movement of the vessel directly by the Android client. This process is achieved by following steps: 1) the Android client sends control commands to the server; 2) the server calculates the steering angle based on the current state of vessel and location information, and sends a corresponding control command to the GPRS module; 3) the GPRS module passes the message to CC2530 chip through the RS485 serial port; 4) the chip simulates PWM waves to control the left and right motor revolution so that vessel can change direction and move freely as expected. The vessel gradually revises its path according to the received data and its current GPS location, and will move towards the final target eventually. Our system has been evaluated in a modern fishery breeding base in Yangzhong, Jiangsu Province. In the experiment, the ship was initially driven by manual control to select twelve measurement positions. After that, we utilize our self-learning system to navigate the ship to access those target positions. The ship stayesat each location for two minutes and collectswater quality parameters in the neighborhood. After an hour of testing, the errors between navigated positions and real target positions areless than 2 meters on average, and the maximum difference of dissolved oxygen value between those positions is 1 mg/L. The change of water temperature is 1.5 °C, and pHvalue remains unchanged. Those results are consistent with the horizontal distribution law of water quality parameters. Compared with current state-of-arts, our system has the capability of mobile data collection, which can not only increase the measurement range but also reduce the cost. The system has significant potential in various applications such as aquaculture, river management, and hydrological monitoring. [ABSTRACT FROM AUTHOR]