Blockbusting and domineering (Q1106742)

We introduce a new game called ``discrete game of approach described by the difference equation (1) \(z(t+1)=Az(t)+\mu (u(t),v(t))\), \(z(0)=z_ 0\), \(t=0,1,\dots\), where \(z\in \mathbb R^ n\), u(t)\(\in U\), v(t)\(\in V\), the sets \(U\) and \(V\) are compact, the function \(\phi(\cdot,\cdot)\) is continuous. The terminal set \(M^*\) is the algebraic sum of a linear subspace \(M^ 0\) of \(\mathbb R^ n\) and a compactum \(M\) from the orthogonal complement of \(M^ 0\) with respect to \(\mathbb R^ n\). The objective of Player 1 who governs the control \(u\) is to ensure that the point \(z(t)\) should contact the set \(M^*\) at some time \(t\). Player 2 who governs the control \(v\) behaves opposed to this. P1 knows the equation and initial condition (1), the sets \(U\), \(V\), \(M^*\) and he possesses complete information not only about the current controls of P2, \(v(t)\), but also about the history of the controls \(v_ t(\cdot)=(v(0),v(1),\dots,v(t))\). The authors give sufficient conditions for successful termination of the game in a finite number of steps. The first theorem which is a discrete version of the first direct Pontryagin's method guarantees the termination of the game in \(P(z_ 0)\) steps by means of a strategy depending only on the current controls \(v(t)\). The second theorem guarantees the termination of the game in \(T(z_ 0)\) steps by means of a strategy depending on the whole history of the controls \(v_ t(\cdot)\). Formulae for \(P(z_ 0)\) and \(T(z_ 0)\) are given. The inequality \(T(z)\leq P(z)\), \(z\in \mathbb R^ n\) holds, in general. Sufficient conditions for the equalities \(T(z)=P(z)\) for all \(z\in \mathbb R^ n\), and \(T(z_ 0)=P(z_ 0)\) for fixed \(z_ 0\) are established.