多传感器观测下带乘性噪声系统的逆向滤波与反褶积融合算法

针对多传感器观测环境下带乘性噪声系统的逆向最优滤波与反褶积融合估计问题 ,本文提出了 1种基于极大似然准则的最优融合算法。该算法中各单传感器间并行计算 ,并且融合中心与单传感器处理中心间无反向通讯 ,因而执行效率较高。仿真表明 ,该融合算法产生的逆向滤波与反褶积比单传感器处理结果有较明显提高     

Waveform Shaping of Sonar Transducers for Improving the Vertical Resolution in Sub-bottom Sediments Profiling

Vertical resolution is of fundamental importance in sonar exploration and is directly related to the duration of the acoustic pulse generated by the transducer. The shorter the radiated pulse, the higher the vertical resolution. Many sub-bottom profiling sonar systems use piezoelectric transducers because they are reversible and well understood. Piezoelectric projectors are normally resonant transducers, which are intrinsically narrowband. A piezoelectric transducer is usually driven by a tone-burst. However, it is possible to use Fourier techniques to find a pre-compensated electrical driving function so that the transducer radiates a prescribed wider band acoustic waveform. This technique can be applied to synthesize zero-phase cosine-magnitude, Gaussian, and bionic pulses, with a conventional sandwich transducer. Zero-phase cosine-magnitude waveforms provide minimum length pulses (and therefore maximum resolution) within a prescribed frequency band.The aim of this paper is to illustrate the synthesis of wideband acoustic pulses using an underwater piezoelectric projector. The conventional acoustic waveform radiated when a Tonpiltz transducer is transiently excited using a “click” and allows its frequency response function to be measured. This function is used to design the electrical signal which then drives the transducer so that it radiates the shortest pulse compatible with its mechanical response. The significant resolution enhancement of the waveform shaping process is illustrated by its application to a sediment wedge model.     

Robust adaptive control of underactuated ships on a linear course with comfort

This paper addresses an important problem in ship control application—the robust stabilization of underactuated ships on a linear course with comfort. Specifically, we develop a multivariable controller to stabilize ocean surface ships without a sway actuator on a linear course and to reduce roll and pitch simultaneously. The controller adapts to unknown parameters of the ship and constant environmental disturbances induced by wave, ocean current and wind. It is also robust to time-varying environmental disturbances, time-varying change in ship parameters and other motions of the ship such as surge and heave. The roll and pitch can be made arbitrarily small while the heading angle and sway are kept to be in reasonably small bounds. The controller development is based on Lyapunov’s direct method and backstepping technique. A Lipschitz continuous projection algorithm is used to update the estimate of the unknown parameters to avoid the parameters’ drift due to time-varying environmental disturbances. Simulations on a full-scale catamaran illustrate the effectiveness of our proposed controller.     

珠江三角洲网河区水位变化趋势研究

根据珠江三角洲网河区29个验潮站的实测验潮记录,应用傅氏变换与最平滤波器串联的方法来消除月均序列的周期波动对确定水位变化趋势的影响,由低通序列一元线性回归分析确定各站水位的变化趋势;结果表明,周期波动对确定海平面变化趋势的影响是显著的.应用经验正交函数(EOF)对网河区的水位变化场进行分解,由相互独立的时间函数和空间特征函数表征网河区区域的水位变化特征;应用时间特征函数计算区域水位的平均变化率为0.02mm/a.根据验潮站的水位变化趋势,探讨网河区水位变化与河床冲淤的关系.     

水下智能潜器的神经网络运动控制

本文介绍一种基于神经网络的水下智能潜器的运动控制方法，该方法通过在线学习，融控制与滤波为一体。计算机仿真与水池实验验证表明，该方法的控制与滤波性能良好，对环境的学习与适应能力强。该方法事实上可用于一般动力系统的控制。     

Growth and Production in Posidonia Oceanica (L.) Delile*

Abstract. In situ investigations of growth and production in a stand of Posidonia oceanica (L.) DELILE at a depth of 4 m at Ischia (Gulf of Naples) were carried out over two growing seasons. Posidonia starts to grow in August and an average bundle produces ten leaves in increasing time intervals until May. Growth curves for the leaves are given. Maximum leaf standing crop is in May with 1300 g dry weight per m^-2, leaf area index at this time reaches 22 m² m^-2. Leaf net productivity is highest in March with 12 g dry weight per m² per day. Annual leaf production is estimated as 3110 g dry weight per m², “underground” production as 115 g dry weight per m². About half the leaf production is exported from the system. Adaptive strategies of the growth and production pattern are discussed.     

Multifractal filtering method for extraction of ocean from remotely sensed imagery

Traditional methods of extracting the ocean wave eddy information from remotely sensed imagery mainly use the edge detection technology such as Canny and Hough operators. However, due to the complexities of ocean eddies and image itself, it is sometimes difficult to successfully detect ocean eddies using these methods. A mnltifractal filtering technology is proposed for extraction of ocean eddies and demonstrated using NASA MODIS, SeaWiFS and NOAA satellite data set in the typical area, such as ocean west boundary current. Results showed that the new method has a superior performance over the traditional methods.     

Dereverberation of marine reflection seismic data by a spatial combination of predictive deconvolution and velocity filtering

Contamination of seismic reflection records at early times by first-order water reverberations can be especially severe during survey operations over hard and flat sea floors on the continental shelf or in lake environments. A new dereverberation scheme based on two classical techniques — predictive deconvolution and velocity filtering — has been developed to address this problem. The techniques are combined spatially to take advantage of their complementary offset- and time-dependent properties. Stage I of the scheme consists of applying predictive deconvolution at short offset. The data are previously conditioned by a normal moveout correction with the water velocity which restores the periodicity of the reverberations in the offset-time plane and enhances the performance of deconvolution. Stage II of the scheme involves velocity filtering in the common-midpoint domain which is particularly effective at long offset where the moveout difference between primary reflections and reverberations is largest. The dereverberation scheme is well suited for the initial processing of large volumes of data due to the general availability of cost-effective deconvolution and velocity filtering algorithms in seismic processing software packages. Practical implementation issues are illustrated by a field example from the GLIMPCE survey in Lake Superior.Lithoprobe Publication No. 475.