Abstract:This paper considers an optimization problem for a dynamical system whose evolution depends on a collection of binary decision variables. We develop scalable approximation algorithms with provable suboptimality bounds to provide computationally tractable solution methods even when the dimension of the system and the number of the binary variables are large. The proposed method employs a linear approximation of the objective function such that the approximate problem is defined over the feasible space of the binary decision variables, which is a discrete set. To define such a linear approximation, we propose two different variation methods: one uses continuous relaxation of the discrete space and the other uses convex combinations of the vector field and running payoff. The approximate problem is a 0-1 linear program, which can be solved by existing polynomial-time exact or approximation algorithms, and does not require the solution of the dynamical system. Furthermore, we characterize a sufficient condition ensuring the approximate solution has a provable suboptimality bound. We show that this condition can be interpreted as the concavity of the objective function. The performance and utility of the proposed algorithms are demonstrated with the ON/OFF control problems of interdependent refrigeration systems.

Non-smooth optimality conditions for a class of dynamic systems with variable structure

Optimal Control for Hybrid Systems Based on Mixed Dynamic Programming

Optimal Vibration Control of a Class of Nonlinear Stochastic Systems with Markovian Jump

An Effective Simultaneous Approach with Variable Time Nodes for Dynamic Optimization Problems

On Modeling and Optimal Control of a Class of Hybrid Systems

Finite-Time Nonconvex Optimization Using Time-Varying Dynamical Systems

A Stochastic Optimal Time-Delay Control for Nonlinear Structural Systems

Adaptive Modular Control for a Class of Nonlinear Systems with Unknown Time-Varying Parameters

Approximation Algorithms for Optimization of Combinatorial Dynamical Systems

Strong Stationarity Conditions for Optimal Control of Hybrid Systems

Optimal Control Of A Class Of Non-Linear Discrete-Continuous Hybrid Systems

Relationships Between the Maximum Principle and Dynamic Programming for Infinite Dimensional Stochastic Control Systems

Convergence of an Algorithm for Optimal Control of Nonlinear Continuous-Time Systems with Model-Reality Differences

Sufficient Optimality Conditions for Hybrid Systems of Variable Dimension with Intermediate Constraints

Optimal state-delay control in nonlinear dynamic systems

A path integration algorithm for stochastic dynamical systems with multiple non-smooth events

Event-triggered sub-optimal control for two-time-scale systems with unknown dynamics

Time Optimal Control Problems for Some Non-Smooth Systems

Adaptive Dynamic Programming and Optimal Stabilization for Linear Systems with Time-Varying Uncertainty

A stochastic hybrid optimal control with regime switching and recursive cost functional

Optimal Control of Discrete-Time Nonlinear Systems