Prediction of Dissolved Oxygen Content in Aquaculture of Sea Cucumber Using Support Vector Regression
How to predict the content of dissolved oxygen (DO) accurately and timely control the dissolved oxygen content is the key problem of healthy breeding in sea cucumber aquaculture. The prediction model which based on support vector regression (SVR) has been proposed in this paper to solve the DO prediction problem. Based on statistical analysis theory, the support vector regression provides a new insight in time series data prediction. It constructs the regression model to describe the non-linear relation between DO prediction target and its influence factors. The radial basis function is used as the kernel function. The support vector regression process is to find the optimal parameters and support vectors which ensure the error between the prediction value and target value is small to negligible. After the regression process, we get the regression model between DO prediction target and its influence factors. We select the PH, electrical conductivity, water temperature, DO of current time to predict the DO of next time. All these data are collected by the water quality monitoring system. By contrast, we also apply the BP neural network in the prediction. The results have shown that SVR has good ability with high precision, lower deviation, and easy to stable.
Dissolved Oxygen (DO) Prediction Support Vector Regression
Yaoguang Wei Daoliang Li Haijiang Tai Jianqin Wang Qisheng Ding
College of Information and Electrical Engineering, China Agricultural University, Beijing 100083, China
国际会议
南昌
英文
1075-7082
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）