Abstract:
In networked control systems, feedback signals are transmitted through digital communication networks, which makes quantization necessary before the transmission of information. However, quantization error is unavoidable and can degrade the system's performance and even its stability. In order to mitigate performance degradation due to quantization error, we introduce an event-triggering technique into these systems' sampling and control functions. With event-triggering control, micro-processors perform sampling only during certain events, which are often defined as certain error variables that go beyond defined bounds. We propose some conditions to guarantee stability under quantization, which can achieve better (longer) average sampling periods than those by conventional periodic sampling techniques. Moreover, we present some quantization density conditions to guarantee stability during logarithmic quantization. The results of simulations conducted on an example system confirm the effectiveness of the stability conditions, and demonstrate the efficiency of the event-triggering technique.