G - Physics – 06 – F
Patent
G - Physics
06
F
G06F 19/00 (2006.01) G05B 15/00 (2006.01) G06F 17/00 (2006.01) H02J 4/00 (2006.01)
Patent
CA 2479603
Load-Flow computations are performed as a step in real-time operation/control and in on-line/off-line studies of Electrical Power Systems. Methods in this application are the best versions of many simple variants with almost similar performance. Simple variants include any possible hybrid combination of these methods. All of the methods are characterized: 1) In the use of a method of decomposing a network referred to as Suresh's diakoptics. Suresh's diakoptics involves determining a subnetwork for each node involving directly connected nodes referred to as level-1 nodes and their directly connected nodes referred to as level-2 nodes and so on. The level of outward connectivity for local solution of a subnetwork around a given node is to be determined experimentally. This is particularly true for matrix based methods such as Newton-Raphson(NR), BGGB, FSDL and SSDL methods. Subneworks can be solved by any of the known methods including Gauss and Gauss-Seidel methods. In the case of Gauss and Gauss-Seidel methods only one level of outward connectivity around each node is found to be sufficient for the formation of subnetworks equal to the number of nodes. Sometimes it is possible that a subnetwork around a node could be a part of the subnetwork around another node making it redundant requiring local solution of less than (m+k) subnetworks. The local solution of equations of each subnetwork could be iterated for experimentally determined two or more iterations. However, maximum of three iterations were fond to be sufficient. 2) Sequential solution of subnetworks involves updating solution estimate of a subnetwork for immediate use in the solution of the next and subsequent subnetworks. This is referred to as block Gauss-Seidel approach. (steps-6 and - 11 in Sequential Algorithm) 3) Parallel solution involves solution of all subnetworks using available solution estimate at the start of the iteration without intermediate updating of solution estimate. It further involves initializing a vector of dimension equal to the number of nodes with each element value zero, adding solution estimates for a node resulting from different subnetworks in a corresponding vector element, counting the number of additions and storing the average element values as initial available estimate for the next iteration.. This is referred to as block Gauss approach. Because a node could be directly connected to two or more nodes or a part of two or more subnetworks emanating from different nodes, a parallel solution iteration involves adding and taking the average of all the solution estimates or corrections obtained for a node in the parallel solution of subnetworks emanating from different nodes (steps-2 and -11 in parallel algorithm). 4) The parallel solution algorithm affords an opportunity to have simplified parallel computer a server procesor-arrey processors architecture, where each of the array processors communicate only with server processor and commonly shared memory locations and not among themselves.(Fig.6)
Na
Patel Sureshchandra B.
LandOfFree
Sequential and parallel loadflow computation for electrical... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Sequential and parallel loadflow computation for electrical..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Sequential and parallel loadflow computation for electrical... will most certainly appreciate the feedback.
Profile ID: LFCA-PAI-O-1600690