E. De Tuglie

A static optimization approach to assess dynamic available transfer capability

This paper deals with the development of a nonlinear programming methodology for evaluating available transfer capability. The main feature of the approach is the capability to treat static and dynamic security constraints in a unique integrated piece of software. The algorithm has been implemented and tested on an actual power system.

Dynamic Security Corrective Control by UPFCs

This paper deals with the development of a nonlinear programming methodology for evaluating corrective actions to improve the dynamic security of power systems when transient instability is detected. Remedial actions are implemented by exploiting the fast response of unified power flow controllers (UPFCs). The algorithm is implemented and tested on the Italian grid.

Real-time preventive actions for the enhancement of voltage-degraded trajectories

This paper deals with the development of a nonlinear programming methodology for evaluating preventive actions to improve the dynamic security of power systems when poor voltage transients are detected. For the proposed power system modeling, voltage-degraded trajectories arise from both angle and voltage instability on the transient time scale. The rescheduling of the generation improves power system security by ensuring an acceptable voltage transient behavior and constraining the system to be stable. The algorithm is implemented and tested on an actual power system.

Transient security dispatch for the concurrent optimization of plural postulated contingencies

The implementation. of a transient security dispatch (TSD) algorithm for real-time applications is presented. The algorithm, based on the generalized reduced gradient (GRG) method, processes collectively a set of contingencies in order to ensure transient stability for all postulated cases. The massive computation has been decomposed on a cluster of workstations, exploiting the decoupled nature of the GRG method and the benefits of distributed computing. The feasibility of the implementation is shown through tests on a realistic-sized test network.

A Coherency Recognition Based on Structural Decomposition Procedure

The aim of this paper is to present an algorithm for coherency recognition. In the proposed decomposition model, the system structure is explored by analyzing and quantifying the strength of connection among generating units by means of suitable coupling factors. These factors take into account interconnection constraints among dynamic components of the system. In particular, coherency indicators are based on the influence that state variables of generators have on the voltages of all other generators. Test results demonstrated that the coherency recognition obtained with this procedure is independent of fault location and of size of disturbance, since it is based on the very structural characteristics of coupling factors. The practicality of the proposed procedure is demonstrated by numerical examples involving the IEEE 118-bus test system.

Nondiscrinilnatory System Losses Dispatching Policy in a Bilateral Transaction-Based Market

A new method to allocate transmission losses for simultaneous bilateral transactions is proposed. The methodology utilizes a circuit approach of the system in conjunction with a classical power flow. For a given operating point, it is possible to derive system loss expression as a sum of partial terms due to each transaction. These transaction loss components, supplied by slack buses, can turn into costs associated to the respective transactions. Alternatively, in this paper, it is proposed that each transaction provides for its own loss, thus eliminating the need for a balancing mechanism. In this case, the developed methodology evaluates the increase of active power at each transaction generator through loss contributions not arbitrarily assumed but calculated from a developed loss allocation formula. The main advantage of the developed method lies in its simplicity and capability of treating multiple transactions simultaneously.

The Characteristic Ellipsoid Methodology and its Application in Power Systems

The characteristic ellipsoid (CELL) method to monitor dynamic behaviors of a power system is proposed. Multi-dimensional minimum-volume-enclosing characteristic ellipsoids are built using synchronized phasor measurements. System dynamic behaviors are identified by tracking the change rate of the CELLs characteristic indices. Decision tree techniques are used to link the CELLs characteristic indices and the systems dynamic behaviors and to determine types, locations and related information about the dynamic behaviors. The knowledge base of representative transient events is created by offline simulations based on the full Western Electric Coordinating Council (WECC) model. Two case studies demonstrate that the CELL method combined with the decision trees can detect transient events and their features with good accuracy.

PrInCE Lab experimental microgrid Planning and operation issues

This paper aims at describing the experimental Microgrid (μG) built at the Politecnico di Bari within a project funded by the Italian Ministry of Education, University and Research. In particular, the μG will provide a test-bed for research, development and testing devices and components for smart-grid applications. Moreover, it will give the possibility to test new control strategies for the optimal management of internal resources and the connection with the distribution company, in order to ensure the proper operation and the integration with the network in compliance with current connection rules.

PV plants for voltage regulation in distribution networks

This paper deals with the development of a centralized nonlinear auto-adaptive controller able to optimize the network voltage profile by managing the reactive power supplied by PV inverters. The control design is based on a realtime optimization procedure involving the sensitivity theory in conjunction with the Lyapunov function and produces the control laws that must be sent to local controllers of PV-inverters. The derived controller is implemented and tested on a MV distribution network.

An extension of feedback-feedforward technique to input-output feedback linearization

Large scale systems, such as interconnected power systems, need to be described in a more accurate form than in the past allowing the overexploitation of system resources. Not negligible nonlinear phenomena increase computational complexity especially for what control system design concerns. Classical control design, based on linearization techniques has become no more suitable to take into account strong nonlinear effects and unmodeled dynamics, while input-output feedback linearization shows computational difficulties as the order of the system to be linearized increases. In this paper the classical feedback-feedforward technique for linear systems has been applied to a particular cascade connection of an input-output linearized system feeded by a linear system. The proposed approach achieves perfect tracking for the linearized model and an overall reduction in computational effort.