Comparison of Multiclass SVM Classification Methods to Use in a Supportive System for Distance Relay Coordination

This paper aims at evaluating the methods of multiclass support vector machines (SVMs) for effective use in distance relay coordination. Also, it describes a strategy of supportive systems to aid the conventional protection philosophy in combating situations where protection systems have maloperated and/or information is missing and provide selective and secure coordinations. SVMs have considerable potential as zone classifiers of distance relay coordination. This typically requires a multiclass SVM classifier to effectively analyze/build the underlying concept between reach of different zones and the apparent impedance trajectory during fault. Several methods have been proposed for multiclass classification where typically several binary SVM classifiers are combined together. Some authors have extended binary SVM classification to one-step single optimization operation considering all classes at once. In this paper, one-step multiclass classification, one-against-all, and one-against-one multiclass methods are compared for their performance with respect to accuracy, number of iterations, number of support vectors, training, and testing time. The performance analysis of these three methods is presented on three data sets belonging to training and testing patterns of three supportive systems for a region and part of a network, which is an equivalent 526-bus system of the practical Indian Western grid.

An intelligent approach using support vector machines for monitoring and identification of faults on transmission systems

Power system disturbances are often caused by faults on transmission lines. When faults occur in a power system, the protective relays detect the fault and initiate tripping of appropriate circuit breakers, which isolate the affected part from the rest of the power system. Generally Extra High Voltage (EHV) transmission substations in power systems are connected with multiple transmission lines to neighboring substations. In some cases mal-operation of relays can happen under varying operating conditions, because of inappropriate coordination of relay settings. Due to these actions the power system margins for contingencies are decreasing. Hence, power system protective relaying reliability becomes increasingly important. In this paper an approach is presented using Support Vector Machine (SVM) as an intelligent tool for identifying the faulted line that is emanating from a substation and finding the distance from the substation. Results on 24-bus equivalent EHV system, part of Indian southern grid, are presented for illustration purpose. This approach is particularly important to avoid mal-operation of relays following a disturbance in the neighboring line connected to the same substation and assuring secure operation of the power systems.

Prediction of Queue Waiting Times for Metascheduling on Parallel Batch Systems

Prediction of queue waiting times of jobs submitted to production parallel batch systems is important to provide overall estimates to users and can also help meta-schedulers make scheduling decisions. In this work, we have developed a framework for predicting ranges of queue waiting times for jobs by employing multi-class classification of similar jobs in history. Our hierarchical prediction strategy first predicts the point wait time of a job using dynamic k-Nearest Neighbor (kNN) method. It then performs a multi-class classification using Support Vector Machines (SVMs) among all the classes of the jobs. The probabilities given by the SVM for the class predicted using k-NN and its neighboring classes are used to provide a set of ranges of predicted wait times with probabilities. We have used these predictions and probabilities in a meta-scheduling strategy that distributes jobs to different queues/sites in a multi-queue/grid environment for minimizing wait times of the jobs. Experiments with different production supercomputer job traces show that our prediction strategies can give correct predictions for about 77-87% of the jobs, and also result in about 12% improved accuracy when compared to the next best existing method. Experiments with our meta-scheduling strategy using different production and synthetic job traces for various system sizes, partitioning schemes and different workloads, show that the meta-scheduling strategy gives much improved performance when compared to existing scheduling policies by reducing the overall average queue waiting times of the jobs by about 47%.

Hybrid Working Set Algorithm for SVM Learning With a Kernel Coprocessor on FPGA

Support vector machines (SVM) are a popular class of supervised models in machine learning. The associated compute intensive learning algorithm limits their use in real-time applications. This paper presents a fully scalable architecture of a coprocessor, which can compute multiple rows of the kernel matrix in parallel. Further, we propose an extended variant of the popular decomposition technique, sequential minimal optimization, which we call hybrid working set (HWS) algorithm, to effectively utilize the benefits of cached kernel columns and the parallel computational power of the coprocessor. The coprocessor is implemented on Xilinx Virtex 7 field-programmable gate array-based VC707 board and achieves a speedup of upto 25x for kernel computation over single threaded computation on Intel Core i5. An application speedup of upto 15x over software implementation of LIBSVM and speedup of upto 23x over SVMLight is achieved using the HWS algorithm in unison with the coprocessor. The reduction in the number of iterations and sensitivity of the optimization time to variation in cache size using the HWS algorithm are also shown.

ELISA: An Estimated Load Information Scheduling Algorithm for Distributed Computing Systems

In this paper, we present a decentralized dynamic load scheduling/balancing algorithm called ELISA (Estimated Load Information Scheduling Algorithm) for general purpose distributed computing systems. ELISA uses estimated state information based upon periodic exchange of exact state information between neighbouring nodes to perform load scheduling. The primary objective of the algorithm is to cut down on the communication and load transfer overheads by minimizing the frequency of status exchange and by restricting the load transfer and status exchange within the buddy set of a processor. It is shown that the resulting algorithm performs almost as well as a perfect information algorithm and is superior to other load balancing schemes based on the random sharing and Ni-Hwang algorithms. A sensitivity analysis to study the effect of various design parameters on the effectiveness of load balancing is also carried out. Finally, the algorithm's performance is tested on large dimensional hypercubes in the presence of time-varying load arrival process and is shown to perform well in comparison to other algorithms. This makes ELISA a viable and implementable load balancing algorithm for use in general purpose distributed computing systems.

Energetics of rare earth manganese perovskites A1-xA ' xMnO3 (A = La, Nd, Y and A ' = Sr, La) systems

High temperature reaction calorimetry using molten lead berate as solvent has been used to study the thermochemistry of NdMnO3, YMnO3, La1-xSrxMnO3 (with 0 < x < 0.5), and Ln(0.5)Ca(0.5)MnO(3) (with Ln = La, Nd, Y), The enthalpies of formation of these multicomponent oxides from their binary constituents have been calculated from the measured enthalpy of drop solution, The energetic stability of the perovskite depends on the size of the A cation, The enthalpy of formation of YMnO3 (smallest A cation) is more endothermic than those of NdMnO3 and LaMnO3. The energetics of the perovskite also depends on the oxidation state of the B site's ions. In the La1-xSrxMnO3 system, the energetic stability of the structure increases with the Mn4+/Mn3+ ratio, The new values of the enthalpies of oxidations, with reliable standard entropies, were used to plot the phase stability diagram of the lanthanum-manganese-oxygen system in the temperature range 300-1100 K, The LaMnO3/MnO phase boundary evaluated in this study agrees with the one published by Atsumi et nl. calculated from thermogravimetric and conductivity measurements.

Spin populations in one-dimensional alternant spin systems: A theoretical approach

Spin-density maps, deduced from polarized neutron diffraction experiments, for both the pair and chain compounds of the system Mn2+Cu2+ have been reported recently. These results have motivated us to investigate theoretically the spin populations in such alternant mixed-spin systems. In this paper, we report our studies on the one-dimensional ferrimagnetic systems (S-A,S-B)(N) where hi is the number of AB pairs. We have considered all cases in which the spin Sri takes on allowed values in the range I to 7/2 while the spin S-B is held fixed at 1/2. The theoretical studies have been carried out on the isotropic Heisenberg model, using the density matrix renormalization group method. The effect of the magnitude of the larger spin SA On the quantum fluctuations in both A and B sublattices has been studied as a function of the system size N. We have investigated systems with both periodic and open boundary conditions, the latter with a view to understanding end-of-chain effects. The spin populations have been followed as a function of temperature as well as an applied magnetic field. High-magnetic fields are found to lead to interesting re-entrant behavior. The ratio of spin populations P-A-P-B is not sensitive to temperature at low temperatures.

Giant Magnetoresistance and Related Properties of Rare Earth Manganates and Other Oxide Systems

Giant magnetoresistance (GMR), which was until recently confined to magnetic layered and granular materials, as well as doped magnetic semiconductors, occurs in manganate perovskites of the general formula Ln(1-x)A(x)MnO(3) (Ln = rare earth; A = divalent ion). These manganates are ferromagnetic at or above a certain value of x (or Mn4+ content) and become metallic at temperatures below the curie temperature, T-c. GMR is generally a maximum close to T-c or the insulator-metal (I-M) transition temperature, T-im. The T-c and %MR are markedly affected by the size of the A site cation, [r(A)], thereby affording a useful electronic phase diagram when T-c or T-im is plotted against [r(A)]. We discuss GMR and related properties of manganates in polycrystalline, thin-film, and single-crystal forms and point out certain commonalities and correlations. We also examine some unusual features in the electron-transport properties of manganates, in particular charge-ordering effects. Charge ordering is crucially dependent on [r(A)] or the e(g) band width, and the charge-ordered insulating state transforms to a metallic ferromagnetic state on the application of a magnetic field.

Thermopower and nature of the hole-doped states in LaMnO3 and related systems showing giant magnetoresistance

We have measured the thermopower (S) of hole-doped LaMnO3 systems in order to see its dependence on the Mn4+ content as well as to investigate other crucial factors that determine S. We have carried out hole doping (creation of Mn4+ by two distinct means, namely, by the substitution of La by divalent cations such as Ca and Sr and by self-doping without aliovalent substitution). The thermopower is sensitive not only to the hole concentration but also to the process employed for hole doping, which we explain as arising from the differences in the nature of the hole-doped states. We also point out a general trend in the dependence of S on hole concentration at high temperatures (T> T-c), similar to that found in the normal-state thermopower of the cuprates.

Precipitation In Small Systems--II. Mean Field Equations More Effective Than Population Balance

Part I (Manjunath et al., 1994, Chem. Engng Sci. 49, 1451-1463) of this paper showed that the random particle numbers and size distributions in precipitation processes in very small drops obtained by stochastic simulation techniques deviate substantially from the predictions of conventional population balance. The foregoing problem is considered in this paper in terms of a mean field approximation obtained by applying a first-order closure to an unclosed set of mean field equations presented in Part I. The mean field approximation consists of two mutually coupled partial differential equations featuring (i) the probability distribution for residual supersaturation and (ii) the mean number density of particles for each size and supersaturation from which all average properties and fluctuations can be calculated. The mean field equations have been solved by finite difference methods for (i) crystallization and (ii) precipitation of a metal hydroxide both occurring in a single drop of specified initial supersaturation. The results for the average number of particles, average residual supersaturation, the average size distribution, and fluctuations about the average values have been compared with those obtained by stochastic simulation techniques and by population balance. This comparison shows that the mean field predictions are substantially superior to those of population balance as judged by the close proximity of results from the former to those from stochastic simulations. The agreement is excellent for broad initial supersaturations at short times but deteriorates progressively at larger times. For steep initial supersaturation distributions, predictions of the mean field theory are not satisfactory thus calling for higher-order approximations. The merit of the mean field approximation over stochastic simulation lies in its potential to reduce expensive computation times involved in simulation. More effective computational techniques could not only enhance this advantage of the mean field approximation but also make it possible to use higher-order approximations eliminating the constraints under which the stochastic dynamics of the process can be predicted accurately.

Bridged systems via radical cyclization reactions. Stereospecific synthesis of chiral tricyclo[6.2.1.0(1,5)]undecanes

The application of radical-mediated cyclizations and annulations in organic synthesis has grown in importance steadily over the years to reach the present status where they are now routinely used in the strategy-level planning.2 The presence of a quaternary carbon atom is frequently encountered in terpenoid natural products, and it often creates a synthetic challenge when two or more quaternary carbon atoms are present in a contiguous manner.3 Even though creation of a quaternary carbon atom by employing a tertiary radical is very facile, creation of a quaternary carbon atom (or a spiro carbon atom) via radical addition onto a fully substituted olefinic carbon atom is not that common but of synthetic importance. For example, the primary radical derived from the bromide 1 failed to cyclize to generate the two vicinal quaternary carbon atoms and resulted in only the reduced product 2.4 The tricyclic carbon framework tricyclo[6.2.1.01,5]undecane (3) is present in a number of sesquiterpenoids e.g. zizzanes, prelacinanes, etc.5

Ionic transport in (PEG)x LiBr systems

Ionic conductivity in (PEG)(x)LiBr systems is measured using the complex impedance method in the temperature range -20 degrees C to 100 degrees C. For x = 6 and 10, above a certain concentration dependent temperature T-c, a power law fit based on mode coupling theory is seen to better explain the data than the Vogel-Tamman-Fulcher (VTF) expression. Li-7 NMR linewidth measurements indicate two regions of motional narrowing, one attributable to segmental motion and the other to translational diffusion.