Learn More
To synthesize fixed-final-time control-constrained optimal controllers for discrete-time nonlinear control-affine systems, a single neural network (NN)-based controller called the Finite-horizon Single Network Adaptive Critic is developed in this paper. Inputs to the NN are the current system states and the time-to-go, and the network outputs are the(More)
A model-based reinforcement learning algorithm is developed in this paper for fixed-final-time optimal control of nonlinear systems with soft and hard terminal constraints. Convergence of the algorithm, for linear in the weights neural networks, is proved through a novel idea by showing that the training algorithm is a contraction mapping. Once trained, the(More)
Large-scale data centers consume megawatts in power and cost hundreds of millions of dollars to equip. Reducing the energy and cost footprint of servers can therefore have substantial impact. Web, Grid, and cloud servers in particular can be hard to optimize, since they are expected to operate under a wide range of workloads. For our upcoming data center,(More)
—The problem of decentralized control of multi-agent nonlinear systems is solved by introducing the concept of virtual agents to generate reference trajectories to be tracked by the actual agents. The tracking problem as an optimal control problem is formulated in the framework of approximate dynamic programming. Solutions are obtained using 'single network(More)
Formation control of network of multi-agent systems with heterogeneous nonlinear dynamics is formulated as an optimal tracking problem and a decentralized controller is developed using the framework of 'adaptive critics' to solve the optimal control problem. The reference signal is assumed available only in online implementation so its dynamics is(More)