Self-tuning run-time reconfigurable PID controller Digital PID control algorithm is one of the most commonly used algorithms in the control systems area. This algorithm is very well known, it is simple, easily implementable in the computer control systems and most of all its operation is very predictable. Thus PID control has got well known impact on the control system behavior. However, in its simple form the controller have no reconfiguration support. In a case of the controlled system substantial changes (or the whole control environment, in the wider aspect, for example if the disturbances characteristics would change) it is not possible to make the PID controller robust enough. In this paper a new structure of digital PID controller is proposed, where the policy-based computing is used to equip the controller with the ability to adjust it's behavior according to the environmental changes. Application to the electro-oil evaporator which is a part of distillation installation is used to show the new controller structure in operation.

In this paper we present an architecture for run-time reconfiguration of network-enabled ubiquitous devices. The whole idea is based on a policy-based system where the whole decision-making (e.g. anomaly detection-related) logic is provided in a form of an externally loaded policy file. The architecture is verified through real-life implementation on an embedded system whose sensitivity can be easily modified should a need arise in run-time without affecting network device/segment (and thus potentially a number of network services) so that they continue working while the re-configuration process is triggered.