A Compositional Approach to Stochastic Optimal Control with Temporal Logic Specifications
Matanya B. Horowitz, Eric M. Wolff, Richard M. Murray
Submitted, 2014 American Control Conference (ACC)
We introduce an algorithm for the optimal con- trol of stochastic nonlinear systems subject to temporal logic constraints on their behavior. We directly compute on the state space of the system, avoiding the expensive pre-computation of a discrete abstraction. An automaton that corresponds to the temporal logic specification guides the computation of a control policy that maximizes the probability that the system satisfies the specification. This reduces controller synthesis to solving a sequence of stochastic constrained reachability problems. The solution to each reachability problem corresponds to the solution to a corresponding Hamilton-Jacobi-Bellman (HJB) partial differential equation. To increase the efficiency of our approach, we exploit a class of systems where the HJB equation is linear due to structural assumptions on the noise. The linearity of the PDE allows us to pre-compute control policy primitives and then compose them, at essentially zero cost, to satisfy a complex temporal logic specification.
- Conference Paper: http://www.cds.caltech.edu/~murray/preprints/hwm14-acc_s
- Project(s): Boeing