Adaptive Vibration Amplitude Controller for Coriolis Mass Flowmeters Based on WHO-BP Neural Network

Based on traditional Proportional-Intergral (PI) vibration amplitude controllers, the Coriosis Mass Flowmeter (CMF) shows slow response and even stopping vibration under gas-liquid two-phase conditions.To solve this problem, WHO-BP(Wild Horse Optimizer-Back propagation) adaptive controller is proposed in this paper. BP neural network is designed to adaptively adjust the parameters of PI controller. WHO is ued to find the optimal solution of the initial weight matrix of BP neural network to prevent vanishing gradient and exploding gradient. The effectiveness of the WHO-BP adaptive controller for improving the drive performance and measurement performance of CMF is verified by experimental and simulations results under gas-liquid two-phase flow conditions. Finally, not only the vibration amplitude stability and robustness of CMF as the flow pattern switching, but also the measurement accuracy under complex flow conditions are improved by proposed algorithm.