基于Kalman滤波数据融合方法的超声波气体泄漏定位研究

超声波泄漏检测定位系统因其精度高、实时性强等特性得到广泛应用。常见的超声波定位算法主要包括到达时间差定位算法与能量衰减定位算法,但两种算法本身的定位精度均不是特别的高。采用基于扩展Kalman滤波的数据融合方法,将TDOA定位法和ED定位法的计算结果进行特征提取并联合得出系统观测值,融合计算出泄漏点与各换能器之间的距离,最终通过几何方法计算出泄漏点的三维坐标。通过实验证明了该算法与传统算法相比,精度更高,结果误差小于20mm。

Ultrasonic Gas Leak Location Research Based on Kalman Filtering Data Fusion Method

An ultrasonic leakage detection and location system has good application prospects because of its high precision, real-time and other features. Ultrasonic location algorithms that are commonly used include the time-difference location algorithm (TDOA) and the energy decay location algorithm (ED). However, positioning accuracy of those algorithms is not high enough. We adopt the data fusion method based on extended Kalman filtering, and transform the arrival time difference of the TDOA location method into distance difference, and also transform the sound pressure ratio of the ED location method into arrival distance ratio by feature attraction. The distance between leak hole and each transducer is calculated through geometrical method, and the corresponding projection coordinate is obtained. It is verified by experiments that compared with traditional positioning algorithms, the proposed algorithm has higher accuracy and its error is lower than 20 mm.