Penalty function solutions to optimal control problems with general constraints via a dynamic optimisation method (Q1187800)

The paper deals with the following optimal control problem: Minimize the functional \(J(u)=\int^ t_ 0 C(t,x(u(t),t),u(t))dt\) over \(U\), under the conditions \(\dot x(t)=f(t,x(t),u(t))\), \(x(0)=x^ 0\), \(h(t,x,\) \(u)=0\), \(g(t,x,u)\leq 0\), \(t\in[0,T]\), where \(U\) denotes the class of all measurable functions \(u: [0,T]\to\mathbb{R}^ r\) such that \(a_ i\leq u_ i(t)\leq b_ i\), \(i=1,\dots,r\), for all \(t\in [0,T]\), \(\{a_ i\}\) and \(\{b_ i\}\) are given real numbers; \(C: [0,T]\times\mathbb{R}^ n\times U\to\mathbb{R}\), \(h: [0,T]\times\mathbb{R}^ n\times U\to\mathbb{R}^ p\) and \(g: [0,T]\times\mathbb{R}^ n\times U\to\mathbb{R}^ q\) are continuous functions. The problem is reformulated as an unconstrained minimization problem using a penalty function approach and solved with Snyman's algorithm [see the first author, Appl. Math. Modelling 7, No. 3, 216-218 (1983; Zbl 0548.65046)].