DEVELOPMENT OF SIGNAL PROCESSING ULTRASONIC ARRAY-SENSOR

本文研制了一种基于孔径编码的超声阵列传感器，对波前的空间分布信息进行多路复用。我们可以通过改变传输波形来控制信号处理函数。实验结果如下：1；我们构建了数据采集系统。采用P （VDF-TrFE）共聚物薄膜按m -序列构建了A-2ch能谱传感器。在同一薄膜上构造了一个发射机阵列，从而减少了传感器的定位问题。静电密封大大降低了噪声，不需要平均。利用变换矩阵在频域的奇异值分解，提出了一种新的成像算子。校准数据是利用与沿直线扫描的目标相对应的实际输出波形获得的。操作人员可以对点目标进行合理质量的成像。该技术通过少量的数据采集和传感器的小扫描范围提供高质量的图像。平均伪影的相对水平降低了-7dB.3。针对回声定位系统，提出了一种新的选择性聚焦方法。换能器被构造成一个用二进制序列编码的线性数组。该阵列将波前转换为串行信号。通过传输时间反转波形来实现选择性聚焦。这使我们能够用串行传输信号控制聚焦模式。将该技术应用于目标强度最大的自适应聚焦中。通过两次迭代，不需要的信号电平降低到4db，但迭代次数超过两次，该电平就会增加。

We have developed a ultrasonic array-sensor, encoding sperture (EA), which multiplex spatially distributed information of wavefront. We can control the signal processing function by changing transmitted waveforms.Following results are obtained :1. We have constructed the data acquisition system. A-2ch EA transducer was constructed by using P (VDF-TrFE) copolymer films according to an M-sequence. The array was constructed with a transmitter on the same film, then the positioning problem of the transducers are reducod. The electrostatic sealed significantly reduced the noise, and no averaging was needed.2. A new imaging operator was developed by SVD of the transfer matrix in the frequency domain. The calibration data is obtained using actual output waveforms corresponding to a target scanned along lines. A point target can be imaged with reasonable quality by the operator. The proposed technique provided images with high quality by a small number of data acquisitions and by a small scanning range of the transducer. The relative level of averaged artifacts is reduced by -7dB.3. A new selective focusing method is proposed for echolocation system. Transducers are constructed to a liner array coded with a binary sequence. The array converts a wavefront to a serial signal. Selective focusing is realized by transmitting a time-reversal waveforms. This enables us to control a focusing pattern with serial transmitting singal. The technique was applied to the adaptive focusing which select the object of maximum target strength. The unneeded signal level was reduced to-4dB by two iteration, this level, however, was increased with iteration number over two.