Oil holdup prediction of oil–water two phase flow using thermal method based on multiwavelet transform and least squares support vector machine

摘要

Oil holdup of oil–water two phase flow (OWTPF) was measured using thermocouple based on the thermal method. A new model based on least square support vector machines (LSSVM) and multiwavelet transform has been proposed for the first time, which is capable of forecasting oil holdup of oil–water two phase flow. The temperature signal of OWTPF is greatly disturbed by noises from external interference, which results in a limited measurement range of oil holdup. In order to solve the problem, a new signal processing method based on the multiwavelet transform is used. Multiwavelet transform has several scaling functions and corresponding wavelet functions, which can simultaneously achieve orthogonality, symmetry. With ideal performance, noises were removed and actual temperature signal was effectively retained. The fluctuated amplitude signal denoised and total flux of OWTPF were employed as inputs and the oil holdup was used as output of LSSVM model. In order to improve the predictive accuracy and generalization ability of the LSSVM model, a Genetic Arithmetic (GA) has been adopted to determine the optimal parameters of LSSVM model automatically. The experiment results indicate that the performance of LSSVM–GA model outperforms those of artificial neural network (ANN), LSSVM–GA model can be used for estimating the oil holdup of OWTPF with reasonable accuracy.