A generalized Nielsen number and multiplicity results for differential inclusions (Q1962101)

The Nielsen number, a lower bound for the number of fixed points of a map, has been used to establish the existence of multiple solutions to differential equations. This paper extends Nielsen theory so that it can be applied to some types of differential inclusions. For a multivalued function \(f\) of a space \(X\) to itself, let \(\Gamma\) be its graph, that is, the set of pairs \((x,y)\) where \(y\in f(x)\) and let \(p,q:\Gamma\to X\) be the projections on the first and second factors respectively, then a fixed point of \(f\), that is a solution to \(x\in f(x)\), is a coincidence of \(p\) and \(q\). The usual Nielsen coincidence theory is not applicable to this setting because of the quite general nature of \(\Gamma\). The authors are able to develop an appropriate Nielsen coincidence theory under sufficiently strong hypotheses. In particular, their theory requires the use of certain covering spaces whose fibers must be finite and this turns out to be a very severe restriction. However, they are able to demonstrate that if \(X\) is an \(n\)-torus, all their conditions are satisfied provided only that comparably weak hypotheses, of the sort required for the classical multivalued Lefschetz fixed point theory, are satisfied. Moreover, in the torus setting, the Nielsen coincidence number that they introduce in the paper is readily computable because it equals the Nielsen fixed point number of a selfmap of a torus. Although the restriction to tori makes the results extremely specialized from the topological point of view, this is precisely the setting that arises when one wishes to apply Nielsen theory to obtain lower bounds for the number of periodic solutions to systems of differential inclusions involving periodic multivalued functions. Thus, as the authors demonstrate by examples, their Nielsen coincidence number accomplishes the task it was intended for.