The Variable Frequency Data Transmission Technology Based on Artificial Neural Network Applying in Measurement while Drilling System

The current continuous wave mud telemetry measurement while drilling (MWD) systems commonly apply binary phase shift keying (BPSK) modulation data transmission. Its communication system is complex, the data transfer rate is low, and the data transmission is not flexible enough to adapt to the environment. Accordingly, the paper explored to use the M-ary frequency shift keying (MFSK) modulation data transmission mode, to research the training algorithm of the neural network and the variable frequency data transmission model. The data reception bit error rate of different transmission modes would be predicted with well trained network, and then chose the data transmission method which had the highest data transfer rate to send and receive the down-hole data in the allowed range of bit error rate. The data reception bit error rate in different drilling conditions and different transmission modes would be recorded. The neural network would be retrained periodically. Thereby the adaptability of the network would be increased. Test results show that the variable frequency data transmission technology based on neural network can achieve higher data transfer rate, improve job reliability, and enhance the adaptability to the environment.