ure 1 IPR exchanges information and takes local decisions to avoid cascade failure when facing a major disturbance, natural or man-made As Figure 2 shows, the IPR receives sensor data, processes the information, takes decisions and commands flow control devices. The routers are organised in a network managing multiple redundant power paths between producers and consumers. The IPR organisation is based in a Peer-to-Peer system (P2P) or a mesh. Figure 2 Proposed architecture for the IPR Figure 3 A simple switch-based IPR system To gain experience with the new simulation environment we developed the simple model of a 3-bus power system depicted in Figure 5. The system consists of two generators and a single load. Each load is controlled by an IPR for a total of 3 IPRs. We also designed a fault injection module that automatically generates a line-to-ground fault on phase A after a pre-specified amount of time. This is an initial experiment and in the future other types of faults will be added to the model. The fault detection circuitry detects the line-to-ground fault by computing the zero-sequence component of the ‘igure 4 Virtual Test Bed (VTB) simulation of the simple IPR system of Figure 3. The units of the plots to the right are Amps in the vertical axis and time in the horizontal axis. Figure 5 A SimPower model of a 3-bus system with 3 IPR modules Figure 6 Current, in Amps, versus time at indicated points in the example 3-bus system Figure 7 SimPower model of DD(X) shipboard power system The model that we have developed uses a new type of computer controllable bus (large buses towards the lower centre of Figure 7. This bus serves the purpose of the EFCD in the IPR architecture. The internal structure of the controllable bus is shown in Figure 8 and it includes sensor and actuator circuits that will be controlled by the ICCU. We are currently developing the IPR inter-communication subsystem that will enable the implementation of the IPR protocols that will be the topic of the following section. 5.1 Current state of the art Figure 8 SimPower model of 3-phase computer controllable bus Figure 9 Electrical network featuring IPR An IPR-based power delivery system scales much in the same way as a computer network scales. Groups of local IPR form ‘local area power networks’ and are interconnected by border IPRs, which enable the formation of larger networks of networks. Meanwhile, border IPRs are responsible for attempting to contain countermeasures and recovery actions inside local area power networks to avoid failures from cascading across large regions of the system. IPRs view currently existing relay-based load shedding schemes as lower layer countermeasures in a multi-layer power delivery network. otes: Gen n: Generator n, SrcPR: Source Power Router, PPR: Principal Power Router, SnkPR: Sink Power Router Figure 10 Relation between EEDN, IPR location and IPR logical connections 5.6 Negotiation in two phases 5.6.2 Inter-zone negotiation Figure 12 Interior and Border IPR Figure 13 Basic operational relationship of IPR major subsystems Table 1 Reliabilities and failure probabilities of IPR configurations The reliability o f each IPR configuration is lower than the reliability of the breaker alone. We expect these results because the reliability in a series system will be less than the lowest reliability of its components. All our IPR configurations reduce to a series configuration. T of the breaker is he only way that the reliability of IPR can be greater than the reliability if we provide a redundant path to the breaker. Does this mean that it is better to have only the breaker instead of the IPR? We believe not. A breaker will act based on local d IPR, through its communication capabilities, will act based enhancing the system reliability. The classical methods d increase in the reliability of a power system when a Special included. However, it is known that when an SPS, like an I significantly improves system res improves the system reliability (Zhu et al., 1997). To pro on local and o not proper Protection Sc perly capture increase in a power system, we will use the Risk Framewor McCalley et al. (Zhu et al., 1997). In the following, we show assessment for a section of the 179- k Assessment ata, without regard to the system state outside its protection zone. The regional data y capture the heme (SPS) is PR, is properly operating, it ponse following a contingency, and therefore, it the reliability developed by a Summary O f the reliability bus system including an IPR. The system will suffer voltage collapse if lines L76—78 and L78—80 simultaneously have an outage. A solution to prevent system collapse is to install VAR compensators in buses 78 and 75. An IPR can be used to activate these compensators in the event that there is one line outage to prevent the collapse in case the other line has an outage, too. Figure 16 Markov chain These values were obtained from the MTBF, MTTF or the annual failure for each component. The failure rate of the router (A,) was obtained from its MTBF of 9.5 years. The daily rate of the software (A2) was calculated from the assumed reliability of 0.95 and converted to the failure rate per year of 0.05129. Finally, the failure rate of the breaker (A3) was obtained from literature. Figure 17 A self-reconfigurable control system Hybrid dynamical systems theory offers a natural framework to deal with the modelling and fault adaptive control problems in EPDS. Consideral ble research work has been dedicated to the study of various aspects of hybrid systems including modelling, stability analysis and control. Intelligence in the context of hybrid d capability of these systems to adapt and reconfigure themse ynamic systems refers to the ves to significant changes in their operating environment and their own structure. An example of a self-reconfigurable control system is the fault adaptive control system described in Simon et al. (2003). Fault adaptive control systems in particular have been studied in the context of robotics (Ji et al., 2003) and manufacturing automation (Prabhu, 2003). 7.3 DC zonal electric distribution system Figure 19 ONR IPS reference system DC zonal distribution system Simulation studies using the DCZEDS model are being performed to understand how it functions and its limitations. Normal operation and fault scenarios are being evaluated. Examples of potential faults are: Figure 20 SIMULINK™ DCZEDS simulation results — fault in Zone 1 7.4 Implementation of the reconfiguration logic Figure 21 System Truth Table for distribution bus faults Current works continue the simulation studies and we are looking at different methodologies to design the logic control component of the supervisory system. Figure 22 System operating modes for the distribution bus fault example The future of power systems for ships will be based on zonal architectures making them an ideal test bed where potential concepts for self-reconfigurable systems for electronic power distribution systems will be evaluated.
