Research On Evaluation Technology Of Surface Water Quality Based On DSA-ELM Model

Water quality evaluation plays an important role in the detection and management of water resources.There are some common methods for water quality evaluation such as single factor evaluation method?comprehensive pollution index evaluation method,fuzzy comprehensive evaluation method,principal component analysis method,water pollution index method,neural network evaluation method.Neural network evaluation method is widely used in classification and evaluation because of high accuracy of classification,superior capability of parallel distributed processing,distributed storage and learning,good robustness and fault tolerance ability and the characteristics of fully approximating complex nonlinear relations.However,with the deepening of the research,the shortcomings of the traditional neural network model is appeared,such as slow convergence speed,Node redundancy,network complexity,and weight adjustment needs.In this paper,the DSA-ELM model was proposed and applied in the water quality evaluation,and the following work was competed:1??Environmental Quality Standard for Surface Water?(GB-3838-2002)was introduced and the water quality evaluation indexes in the standard was analyzed.After introducing the detection methods of several main indexes,a detection circuit of pH value based on pH sensor and single chip microcomputer is designed and manufactured.A convenient and effective detection circuit for water quality index detection was provided.2?According to the predation and life habits of dolphins swarm,an optimized bionic swarm intelligence algorithm-Dolphins Swarm Algorithm(DSA)was proposed in this paper.The basic idea and algorithm steps of DSA were introduced,and the superiority was verified by comparing with the traditional particle swarm algorithm.In view of the shortage of the traditional extreme learning machine(ELM)model,a new model-Extreme Learning Machine optimized by Dolphins Swarm Algorithm(DSA-ELM)was proposed.The optimal connection weights and thresholds of the model are searched by Dolphin Swarm Algorithm,so that the new model can achieve optimum only by setting fewer hidden layer nodes.3?The DSA-ELM model was used in the evaluation of water quality.According to regulation of water quality evaluation index in ?Environmental Quality Standard for Surface Water?(GB-3838-2002)and ministry of environmental protection of China,four factors of Huaihe River including the pH value,dissolved oxygen,permanganate index ammonia nitrogen index were chosen to evaluate water and DSA-ELM model was used to evaluate the water quality of 18 points in Huaihe river.The practicability and superiority of DSA-ELM model for water quality evaluation are proved by comparing with the PSO-ELM and ELM model.Water quality of 3 points in 20 weeks was analyzed through DSA-ELM model and it is confirmed that DSA-ELM is a new superior model for water quality evaluation.4?The interface system of water quality evaluation result was designed and manufactured by LabVIEW software which include user login system and water quality evaluation display.Water quality evaluation was realized by MATLAB script node of LabVIEW and matlab mixed programming.By choosing the week the user can observe the water quality data and the evaluation result in display interface as well as the chart display.The water quality limit alarm is designed.The user also can observe all the water quality data and the evaluation in the form.This study shows that DSA-ELM model has the advantage of high accuracy,good stability and short running time and it is feasible and superior in the actual application of water quality evaluation.The results of the evaluation of water quality can be displayed by LabVIEW interface.A new method was provided for the development of water quality evaluation technology in this paper.