Water Quality Evaluation Model Based on Principal Component and Particle Swarm Optimization Support Vector Machine

Abstract

Water is an important natural resource and an essential material guarantee for human survival and development. With the continuous development and progress of social economy, people's demand for water resources is higher and higher, but the current water environment is not optimistic, so the water quality evaluation and management is very urgent. In practical work, in order to ensure that the water quality reaches the standard, a variety of methods need to be used for testing, the most common of which is water sample analysis. However, due to the immature testing technology, and most of the tasks of water sample determination are completed manually, many samples are difficult to obtain accurately, resulting in large deviation of the test results, which cannot meet the current water quality monitoring requirements. Principal component analysis, particle swarm optimization (PSO) and support vector machine (SVM) models are widely used in the research field of water pollutant content as effective prediction and analysis methods, which can well link various elements in the water environment and realize reasonable judgment and control of the impact laws and change laws of different substance concentration changes. At the same time, it plays a very positive role in protecting the ecological environment. In this paper, the principal component analysis model and PSO algorithm were described in detail. Combining the SVM model and the adaptive learning mechanism of PSO, a set of online water environment quality monitoring system based on PSO was designed and developed, which provided technical support for solving the problems of the lack of representativeness and poor accuracy of traditional water quality monitoring data. Compared with the traditional single neural network, the result showed that the SVM-based evaluation model had a higher prediction effect, and the accuracy was also improved by about 6.5%, which can reflect the water pollution situation more truly and reliably.