A Kind Of Vibratory Isolation Algorithms Based On Neural Network

Vibroseis is a green environment-friendly exploration equipment,it can produce frequency continuous changes and constant amplitude of seismic excitation signal,so it in the field of oil exploration has been widely used,but at present application of vibrator of a vibration isolation system is usually a passive vibration isolation system can be produced by the engine and hydraulic system of interference signal isolation,but when the signal is lower than 6Hz,the passive isolation system is difficult to meet the requirements of better isolation.The low frequency interference signal will be external signal effectively buried,resulting in unable to detect some important geological structure.However,due to the low frequency signal penetration depth,low frequency exploration will become a development trend in the field of petroleum exploration.In this paper,a hybrid vibration isolation system is used to improve the vibration isolation efficiency of the low frequency interference.The main work completed includes:Firstly,to improve interference of low frequency vibration isolation efficiency the research purpose,the preliminary work is clearly the vibroseis system structure and working principle,based on established vibroseis across the physical model and mathematical model of the vibration system.Furthermore,passive vibration isolation system and the hybrid vibration isolation system of vibration isolation mechanism are analyzed,for the subsequent to build a simulation system to do the groundwork.Secondly,by comparing the advantages and disadvantages of PID control algorithm,adaptive filter feedforward control algorithm,fuzzy control algorithm and neural network predictive control algorithm,and their respective application conditions.Due to the neural network predictive control algorithm application of the information in the past,now and in the future information has the forecast function,and has strong fault-tolerant ability and robust performance.The neural network predictive control algorithm is applied in the end.Then,a neural network predictive control model is generated in Simulink software,and the parameters of the model are modified and debugged repeatedly,until the vibration isolation system has the best control effect.Thirdly,according to the mathematical model of the hybrid vibration isolation system,a complete simulation model of the system is built by using Simulink software package.The simulation results are compared and analyzed.The results show that the vibration isolation algorithm based on neural network has a very obvious control effect on the existence of nonlinear active vibration isolation system.When the frequency of the excitation signal is 6Hz,hybrid vibration isolation system of vibration isolation efficiency can reach 96%,at 3 Hz,the vibration isolation efficiency reached 92%,visible application of neural network prediction control algorithm for mixed vibration isolation system on low frequency disturbances still has obvious effect of vibration isolation.