We spend a significant amount of time by artificial lighting. This is the reason why the intelligent lighting control has a big role in everyday life. These systems save energy, because they can adjust the intensity of the light sources to the ambient light conditions. There can be more light sources and several different light conditions in the same room, which we must consider while measuring. Therefore this is a MIMO (Multiple Input Multiple Output) system. The outer light depends on many factors, it can change suddenly, for example if a cloud hides the sun, hence the system has a lot of uncertainties. For this reason, the robustness of the controlling algorithm is especially important.
The goal of my dissertation is to plan and build this system at home. The control is performed by a negative feedback loop, like a PID controller, which is really useful in the industry. In the system there are crossovers, one output does not only depend on one input. So while solving the problem, we have to define these interactions and separate the control signals. After the separation, one output only depends on one input. As we control these, we can calculate the original control signals.
Besides the solutions found in literature, I try to control the outer light with an automatic blinder, that can be controlled by this system. Furthermore, I present an optimization algorithm, which minimizes the error in the LS sense and saves energy by preferring the outer light to the artificial lights.