松动件定位外文翻译--松动件位置的技术研究(编辑修改稿)

松动件定位外文翻译--松动件位置的技术研究(编辑修改稿)内容摘要：

ssary to identify the location of loose parts and to remove them from the RCS. Due to this necessity, much research has beencarried out for the development of a method to estimate the location of the loose and Shahrokhi (1981) introduced a method to estimate the impact location using both the arrival time difference and the damping value of impact signals measured with two accelerometers. Olma (1985)also introduced a method by using both the arrival time difference between longitudinal and transverse waves measured with one accelerometer and the propagation velocity of the two waves. However, the estimation of the arrival time difference in these methods based on only the shape of time history of impact signals result in sometimes incorrect estimation of the arrival time difference. The method proposed in this paper uses dispersion characteristics of bending waves in a plate, where propagation velocity varies with frequencies. The dispersion characteristics of bending waves can be fund through the transformation of impact signals using the WignerVille distribution. The arrival time difference can be obtained more accurately using the dispersion characteristics of bending waves. The distance from the impact location to the signal measuring point can be estimated more accurately usingthe information on the power propagation velocity and the arrival time difference of two bending waves. The experimental results show that the proposed method estimates the impact location with relative percentage error within 10% pared with the actual impact location. 2. WIGNERVILLE DISTRIBUTION As one of the methods enabling simultaneous signal analysis in time and frequency domain, the WignerVille distribution has been drawing a lot of attention lately. The WignerVille distribution was first proposed by Wigner (1932), and the concept of the distribution was reestablished by Vitle (Ville, 1948).The WignerVille distribution, W(t, co ) is defined as   detw j   / 2 ) .(t*2 ) . s/s ( t),( （ 1） where t, ɷ,τ , , s(t) represents the time, the angular frequency, the time delay and the time history respectively, and the asterisk (*) denotes the plex conjugate. From Eq.（ 1）the WignerVille distribution is regarded as Fourier transform for the time delay r of s(t +τ / 2 ) . s * ( t τ / 2 ) which is a timedependent autocorrelation function, and represents the distribution of power on the timefrequency order to calculate the WignerVille distribution from a signaldata of limited record length, an approximate discrete value of Eq. (1) must be calculated. The approximate discrete value of Eq. (1) is suggested by . Kim and . Park (1994) as KnkiNNk m egtntm/21 )k(2,(W   （ 2） where ])[(].)[()( * tkmStkmskg m  and k, m and n are integers and N is the number of time history data. Also, K=2N,∆ɷ =π/k∆t and ∆t is the sampling time. Using the periodicity of signal data with limited record length as suggested by Wahl and Bolton (1993), Eq. (2) can be expressed as) KnkiNNk m ekgtntmW/21 )(2),(   （ 3） It can be seen from Eq. (3) that W（ m∆t,n∆ɷ) is related to the discrete Fourier transform (DFT) of gm(k). 3. DISPERSION CHARACTERISTICS OF BENDING WAVES IN A FLAT PLATE The equation of bending wave propagation on a flat plate, showing dispersion characteristics, can be derived from the wave。