College of Aerospace Engineering, Chongqing University, Chongqing 400044, China.
State Key Laboratory of Ultra-precision Machining Technology, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong.
Ultrasonics. 2023 Jul;132:106982. doi: 10.1016/j.ultras.2023.106982. Epub 2023 Mar 8.
In this paper, the formation of theoretical error is presented to investigate the acoustic source localization (ASL) error that can be expected from traditional L-shaped, cross-shaped, square-shaped, and modified square-shaped sensor cluster arrangements. The response surface model based on the optimal Latin hypercube design is developed to theoretically study the effects of sensor placement parameters on the error evaluation index of root mean squared relative error (RMSRE) for the four techniques. The ASL results from the four techniques with the optimal placement parameters are analyzed theoretically. The relevant experiments are conducted for verifying the above theoretical research. The results show that the theoretical error, formed by the difference between the true and the predicted wave propagation directions is related to arrangement of sensors. The results also show that the sensor spacing and the cluster spacing are the two parameters that affect the ASL error most. Between these two parameters the sensor spacing has the stronger influence. The RMSRE increases with an increasing sensor spacing and a decreasing cluster spacing. Meanwhile, the interaction effect of placement parameters should be also emphasized, especially that between the sensor spacing and the cluster spacing for the L-shaped sensor cluster-based technique. Among the four cluster-based techniques, the newly modified square-shaped sensor cluster-based technique shows the smallest RMSRE and not the largest number of sensors. This research on error generation and analysis will provide guidance for the optimal sensor arrangements in cluster-based techniques.
本文提出了理论误差的形成，以研究传统 L 形、十字形、方形和改进方形传感器组排列可能产生的声源定位（ASL）误差。基于最优拉丁超立方设计的响应面模型被开发出来，从理论上研究传感器放置参数对四种技术的均方根相对误差（RMSRE）误差评估指标的影响。从理论上分析了四种技术在最优放置参数下的 ASL 结果。相关实验验证了上述理论研究。结果表明，由真实和预测波传播方向之间的差异形成的理论误差与传感器的排列有关。结果还表明，传感器间距和簇间距是影响 ASL 误差的两个最重要的参数。在这两个参数中，传感器间距的影响更大。RMSRE 随着传感器间距的增加和簇间距的减小而增加。同时，还应强调放置参数的交互作用，特别是对于基于 L 形传感器簇的技术，传感器间距和簇间距之间的交互作用。在基于簇的四种技术中，新改进的基于方形传感器簇的技术显示出最小的 RMSRE，而不是最大数量的传感器。这项关于误差产生和分析的研究将为基于簇的技术中的最佳传感器布置提供指导。
