Research On Short-term Forecast Method Of Algal Bloom Based On Mobile Detection Platform

Nowadays, the algal bloom is becoming one of the most severe water environmental prob-lems. The algal bloom not only destroys the water environment, but also harms the health of human beings. Unfortunately, there are few effective measurement ways to solve this problem in the short term. Therefore, it is of great significance to forecast the timely algae concentration accurately, reduce the protection cost, and take emergency measures before the algal bloom.A segmented short-term forecasting model based on the generating mechanism of the algal bloom had been set up based on the literature study. The mobile detection system was designed and improved which is cost-efficient, portable, convenient for operation, and feasible for online water quality parameter detection. By combining the segmented short-term forecasting model and mobile water quality detection system, the real time algal bloom in different surface water areas can be forecasted. The main contents and innovations of this thesis are as follows:(1) Segmented short-term forecasting model was set up based on the generating mechanism of algal blooms. Firstly, key factors of algal blooms are determined, such as temperature, light intensity, nutrient salt. Secondly, according to the essential generating process of algal bloom, the growth rhythm model of algal bloom was completed. Thirdly, the model was developed with the characteristics of segmented generating mechanism based on four phases theory of the algal blooming.(2) The stimulation experiment of the model was completed by using the data of Elbe River from March to October in 2000. The parameters were calibrated dynamically with the Particle Swarm Optimizer. Moreover, the feasibility of the model was analysised for forecasting chlo-rophyll-a concentration from four aspects, which are, the time span of calibration data, the group of calibration parameters, forecasting series of the dynamic parameter-calibration, and the forecasts in three days. The results are:a) the forecasting results are the best when the time span of calibration data was 7 days. b) The calibration of the light semi saturation constant has more effects than the optical coefficiency. c) The error of the mechanism forecasting series is approximately 10%, so the accuracy of the forecast is high. d) The errors of forecasts are in-creasing from the first day to the third day.(3) The hardware and software of mobile water quality detection system had been designed and improved in this thesis, which is made up of a mobile detection platform of water quality parameters, a monitoring center and a handheld terminal. The detection platform is used to de-tect the water quality in target water areas, and send the results to the monitoring center and the handheld terminal. Remote detections of temperature and light intensity are feasible. Monitor-ing center is used to store the water quality information and achieve the forecasts of algal blooms. Handheld terminal is used to control the detection platform. The improved system can detect a varity of water quality parameters simultaneously, acquire the color feature of the im-ages to quickly recognize the algal blooming by color sensor on the platform.