Search

Hybrid control problems are complicated by the need to make a suitable sequence of discrete decisions related to future modes of operation of the system. Model predictive control (MPC) encodes a finite-horizon truncation of such problems as a mixed-integer program, and then imposes a cost and/or constraints on the terminal state intended to reflect all post-horizon behaviour. However, these are often ad hoc choices tuned by hand after empirically observing performance. We present a learning method that sidesteps this problem, in which the so-called N-step Q-function of the problem is approximated from below, using Benders’ decomposition. The function takes a state and a sequence of N control decisions as arguments, and therefore extends the traditional notion of a Q-function from reinforcement learning. After learning it from a training process exploring the state-input space, we use it in place of the usual MPC objective. We take an example hybrid control task and show that it can be completed successfully with a shorter planning horizon than conventional hybrid MPC thanks to our proposed method. Furthermore, we report that Q-functions trained with long horizons can be truncated to a shorter horizon for online use, yielding simpler control laws with apparently little loss of performance., Proceedings of Machine Learning Research, 120, ISSN:2640-3498, Proceedings of the 2nd Conference on Learning for Dynamics and Control

This is our prototype for a smart switchboard in the home automation domain. Unlike normal switchboards, each socket on the switchboard is designated as a function (light, fan, and power) and is controlled by a microcontroller, which receives inputs from analog temperature and light sensors. This smart switchboard helps in collecting data about the surroundings, and stores and analyzes the data to identify optimal lighting and cooling pattern for energy efficiency. The smart switchboard can be controlled remotely or manually. It is embedded with sensors and has the ability to communicate with the Internet and upload data to the Internet, which makes it an Internet of Things object.