Two, more readily computable equivariant Nielsen numbers I. Nielsen theory for \(\mathbf M\)-ads (Q2378285)

A space \(X\) is said to be \(M\)-ad if there is a collection \(\{X_v\}_{v\in M}\) of subspaces of \(X\) which are indexed by a set \(M\). A map between two \(M\)-ad spaces \(X\) and \(Y\) having the same index set \(M\) is said to be \(M\)-ad if \(f(X_v)\subset Y_v\) for all \(v\in M\). This paper deals with the estimation of coincidence points of \(M\)-ad maps. Resorting to the method in relative Nielsen fixed point theory, two invariants \(N(f, k; X -\{X_v\}_{v\in M} )\) and \(N(f, k; X, \{X_v\}_{v\in M} )\) are defined for two \(M\)-ad maps \(f\) and \(k\) with domain \(X\). The first one is a generalization of the Nielsen number of a relative map on the complement introduced by the reviewer [Lect. Notes Math. 1411, 189--199 (1989; Zbl 0689.55008)], and gives a lower bound of the number of coincidence points on the complement \(X-\bigcup_{v\in M} X_v\) of all maps \(f'\) and \(k'\) which are homotopic as maps of \(M\)-ads to \(f\) and \(k\) respectively. The second one is a generalization of the primary relative Nielsen number of \textit{H. Schirmer} [Pac. J. Math. 122, 459--473 (1986; Zbl 0553.55001)], and gives a lower bound of the number of coincidence points on all of \(X\). The concept of \(M\)-ads is an abstraction of equivariant maps of between equivariant spaces. This is the reason why the title contains the word ``equivariant'', which is not really used in this paper. In fact, the idea here will be applied to define some Nielsen type numbers for equivariant maps in the author's next paper.