Research On Wave Measurement Method Based On Inertial Measurement Elements

With the development and utilization of marine resources,wave measurement in the ocean has become a key research object.In this paper,based on some problems existing in the wave measurement of current laboratory,the application of inertial measurement components in wave measurement engineering is proposed.The main contents of this paper are as follows:In this paper,the working principle and output mode of the inertial measuring element accelerometer and gyroscope have been studied,and the attitude angle of the gyroscope output of the inertial measuring element has been combined with the theory of the inertial navigation system,and the transformation matrix model in the commonly used coordinate system,that is,the attitude matrix,has been established;The attitude matrix has been used to transform the acceleration output of the inertial measurement component accelerometer from the carrier coordinate system to the geographic coordinate system required for the experiment.The algorithm of the attitude matrix has been studied: the direction cosine method,the quaternion method and the euler angle method;The glitch and DC components in the original acceleration signal have been removed by the five-point three-time moving average method and the mean value method respectively.The Simpson formula in the time domain integration algorithm has been used to integrate the acceleration signals after removing the glitch and DC components;The elimination algorithm of polynomial trend items have been studied,that is,least squares,sliding averaging method and digital filtering method,and have eliminated the polynomial trend items in the integral process by using three algorithms,and the polynomial trend item elimination effect of three kinds of algorithms have been compared and analyzed;The test results show that the three algorithms can make the vertical displacement curve of the strapdown wave height measuring instrument agree well with the measured liquid level curve of the capacitive wave height sensor;From the average period relative error analysis,the average period relative error of the two instruments is almost zero,indicating that the strapdown wave height measuring instrument has high wave period measurement accuracy;From the average wave height relative error and the average peak error analysis,the average wave height relative error and the average peak error of the moving average method are smaller than the least squares method and the digital filtering method,which indicates that the sliding average method makes the wave height measurement accuracy of the strapdown wave height measuring instrument higher;From the average wave height relative error analysis,for the same period,different wave height conditions,the average wave height relative error of the two instruments increases with the increase of the wave height,for the same wave height,different period,the average wave height relative error of the two instruments increases with the increase of the period.This test proves the feasibility of the wave measurement method proposed in this paper for wave period and wave height measurement.The variation process of the wave from the peak to the trough has been studied and thefrequency of the XY direction velocity at the zero point has been calculated at the frequency of each angle interval.The angle range with the highest frequency is the wave direction.The test results show that the frequency of the clip in the XY direction is the highest in the ESE~SE angle range,and the wave direction is in the ESE~SE angle range,which is consistent with the test at that time.This test proves the feasibility of the wave measurement method proposed in this paper for wave wave direction measurement.The structural characteristics and learning algorithm of Elman neural network in artificial neural network have been studied.The roll angle amplitude,the pitch angle amplitude,the heading angle average,the amplitude of the sky acceleration in the geographic coordinate system,and the period of the wave have been taken as the neurons of the input layer,and the wave height has been used as the neurons of the output layer.The wave height prediction model of the Elman neural network has been established,and the simulation and result analysis of the wave height prediction model of the Elman neural network have been performed.The experimental results show that the wave height prediction of Elman neural network can reach the expected error of 0.01;From the average relative error analysis of wave height,the average relative error of Elman neural network is 5.2%,which proves the feasibility of Elman neural network for wave height prediction.