Av. Knyazev et Al. Skorokhodov, PRECONDITIONED GRADIENT-TYPE ITERATIVE METHODS IN A SUBSPACE FOR PARTIAL GENERALIZED SYMMETRICAL EIGENVALUE PROBLEMS, SIAM journal on numerical analysis, 31(4), 1994, pp. 1226-1239
It is shown that a modification of preconditioned gradient-type iterat
ive methods for partial generalized eigenvalue problems makes it possi
ble to implement them in a subspace. The authors propose such methods
and estimate their convergence rate. The authors also describe iterati
ve methods for finding a group of eigenvalues, propose preconditioners
, suggest a practical way of computing the initial guess, and consider
a model example. These methods are most effective for finding minimal
eigenvalues of simple discretizations of elliptic operators with piec
ewise constant coefficients in domains composed of rectangles or paral
lelepipeds. The iterative process is carried out on the interfaces bet
ween the subdomains. Its rate of convergence does not decrease when th
e mesh gets finer, and each iteration has a quite modest cost. This pr
ocess is effective and parallelizable.