Algal Growth Model And Prediction Based On Sparkr

The identification and modeling prediction of algal growth influencing factors has always been the research focus of algal bloom prevention and treatment.The upcoming big data era of water quality index led by the development of water quality monitoring and automation requires a data analysis platform that combines big data with traditional analysis.The evolution of complex systems for algae growth requires an intuitive description.In response to these problems,the paper has done the following research:(1)By deploying a cluster environment on a virtual machine by combining multiple components of the big data ecosystem,the SparkR water quality analysis platform is built with MySQL+Hive+SparkR as the main framework,which implements a whole set of processes from data input,storage,scheduling to application analysis.Big data technology is combined with R language analysis to meet the requirements of reliable storage,unified scheduling and data volume scalability analysis of water quality monitoring index data.At the same time,ECHarts technology is used to realize the visual display of water quality data.(2)In order to avoid the interference of environmental and human factors,the experimental group and the repeated experimental group of indoor simulation culture were designed to simulate the grass and algae artificial lake environment indoors,while the indicators of the control group without the planting grass and the bitter grass group with the bitter grass were monitored.The data were processed and dispatched by SparkR platform.The results were analyzed by algae growth process and correlation analysis.The main influencing factors of algae growth in control group and Valeriana vulgaris group were identified by using Adaptive-Lasso algorithm,and the adaptive-Lasso regression equation was established to verify the results.The prediction model of algae was established by using GBRT algorithm in cluster environment.Through comparative analysis,it is proved that Valeriana vulgaris can remove phosphorus,nitrogen and inhibit algae in water environment,while the main influencing factors of algae in control group and Valeriana vulgaris group are pH,dissolved oxygen,turbidity,conductivity,total phosphorus,total nitrogen and pH,dissolved oxygen,turbidity,conductivity,total phosphorus and total nitrogen respectively.Repeated experiments showed that the relative error of GBRT algae prediction models in the next three days was 15.3%and 14.8%respectively.(3)The EUTRO eutrophication module of WASP model was used to construct the algae growth model.Combined with the experimental data,the algae growth,ammonia nitrogen,total phosphorus and dissolved oxygen modules were selected to construct the model and calibrate the parameters,while the corresponding model was tested.With algae energy value as attribute index,algae is abstracted as agent and its behavior rules are defined based on model and algae activity behavior,while ammonia nitrogen,total phosphorus and dissolved oxygen are abstracted as the main body,and multi-agent model is constructed by combining multi-agent layer and environment layer.Finally,the multi-agent model of algae growth is simulated by NetLogo platform.