scientific article; zbMATH DE number 17520

From MaRDI portal
Publication:3974844





[For the entire collection see Zbl 0697.00033.] In the paper the following practical problem is discussed. In the plane \(x0y\) there is an acoustic scalar receiver located at the point \((0,R)\). On the \(0x\) axis \(n\) acoustic sources are moving following the law \(x_ k(t)=v(t-\tau_ k)\), \(y_ k\) is equal to zero and \(r_ k\) denotes the point \((x_ k,0)\). The contribution of the \(k\)th moving source at the receiving point is assumed to be given by \(s_ k(t)=p_ kA(d_ k)q_ k(t)\), where \(p_ k\) is the source intensity, \(d_ k\) is the distance between the receiver and the point \(r_ k\), \(A(u)=1\sqrt u\) and \(q_ k(t)\) is a multiplicative noise corrupting the \(k\)-th signal. Therefore the total received signal is \(s(t)=\sum^ n_ 1 s_ k(t)\). Now, if one observes the received signal over a finite length interval \([0,T]\), it is asked to find the (unknown) number \(n\) of sources. It is shown that the so formulated problem can be converted in that of solving a Fredholm integral equation of the first kind. After giving some theoretical results, an algorithm to obtain \(n\) is presented. A numerical experiment confirms the validity of the proposed method.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3974844)