Research On Ground Alarm And Positioning System Of Random Acoustic Array

Passive acoustic detection technique has been commonly used in the military field by reconnaissance and defensive measures, which have the advantages of good concealment and far transmission distance etc. In the existing research on the passive acoustic detection technique, there is only target localization based on fixed formation, such as linear array, circular array and four element array etc. But the target localization based on Wireless Acoustic Sensor Network is researched rarely. Wireless Acoustic Sensor Network includes a certain number of wireless microphone nodes, which can be can be freely arranged in the space and forms self-organizing network. Random acoustic array which is researched in this issue is more convenient in deployment, covers a larger area than fixed array, and it’s array size is not limited. So signal processing theory, theoretical modeling, node-deployment which are contained in Random acoustic array has become a hot research in the field of passive acoustic detection.The array positioning technology is applied into wireless sensor network, and the target sound source is calculated by using four node model based on the target localization method of the sound energy. The unite-linear least squares algorithm is improved, and the attenuation coefficient is introduced into the algorithm as a variable. At the same time, the new algorithm abandons the way from original algorithm by increasing the number of nodes to improve the positioning precision. The influence on positioning error is analyzed by the shape, size and position between each other of the random four node model and the optimal four node array which makes the positioning error minimum is found out through a lot of simulation data. Meanwhile, the system design of random four node array is completed, and the four node positioning system is used to measure and calculate the signal, which verifies the positioning effect of the four node optimal model proposed in this paper. The test results show that the optimal model is better than the other models in the positioning effect.