Signal Processing Of NQR Explosive Detection System Based On TMS320C6713

In recent years, all kinds of terrorisms are gradually prevailing in the world because of the changes of international situation. In order to reduce the occurrence of the terrible events, all countries are promoting their technologies to make sure safety. It is especially important to detect explosives in the field of the security inspection. A new method, Nuclear Quadrupole Resonance (NQR), is able to detect whether there are explosives in a packet, easily and exactly. Many experts in the word are attaching importance to it, and have worked much on it.All kinds of explosives contain a chemical element, 14N. This characteristic can be used for Nuclear Quadrupole Resonance to detect explosives. This method not only tells the existing of explosives, but also the sort. The Nuclear Quadrupole Resonance signal is rather weak, and it is hard to extract it from background interference and noise, which is limiting the application of the technology. Some difficult problems about the processing of Nuclear Quadrupole Resonance signal must be solved before the technology is applied to detect explosives in practice. Based on the principles of Nuclear Quadrupole Resonance, this article puts forward some useful methods about signal processing according to the characteristic of NQR signal.There is very serious interference (so-called ringing), an adaptive filter is designed to kill the ringing effectively. The signal-to-noise ratio of NQR signal is remarkably increased.Leakage of major lobe and interference between consecutive spectrums come into being when a long sequence is cut down in the course of DFT, which reduced the accuracy of signal processing. A method of classic power spectral estimation is applied to eliminate these leakage and interference; as a result, the accuracy of detecting explosives is improved greatly.A high performance DSP chip TMS320C6713 is used as a platform to accomplish the complicated algorithm of signal processing. It has proven that background noise is effectively cut down, and serious interference is killed. The aim of researching a sample machine of detecting explosives is accomplished, which is a good effort to apply NQR theory to practice.