A LSTM based method to remove the time-varying errors for mm-wave microfluidic measurement

Time-varying errors bring adverse effects to the mmwave microfluidic measurement system. In this paper, a method based on LSTM model is introduced to predict the drift errors, hence the consistency can be improved and the cost of time can be reduced, especially in mm-wave band. The paper started with analyzing of drift errors source then used LSTM model to build prediction model for drift errors. Experimental result showed that this method can predict drift errors effectively and improve the reliability of long-term measurement at mm-wave frequencies.