Practical method for generation expansion planning based on dynamic programming

This paper presents an efficient method which provides the optimal generation mix and the optimal generation construction process. The approximation method in which the dynamic programming technique and gradient method are combined is applied to determine the optimal generation mix with hydropower generation technologies. The successive approximations dynamic programming (SADP) technique, which is very suitable for high-dimensional multistage decision process problems, is used for obtaining the optimal generation construction process. The effectiveness and feasibility of the developed technique are demonstrated on a practical power system model which has five types of generation technologies including a hydropower generation technology.     

Thermomechanical stress fields in high-temperature fibrous composites. I: Unidirectional laminates

This two-part paper presents a closed-form procedure for evaluation of estimates of local thermomechanical stress fields in two-phase fibrous composites and laminates. The first part is concerned with a unidirectional elastic laminate subjected to uniform mechanical loads and to uniform changes in temperature. Both phases are assumed to be elastic, with temperature-dependent moduli and expansion coefficients; the solution reflects the influence of thermomechanical interactions. Exact solutions are not possible for any real system, because the local geometry is not known in detail. Instead, estimates of the fields are found from a modified Mori-Tanaka approximation. Examples are presented for two SiC/Ti---Al---Nb composites. Local stresses of interest are found after cooling from fabrication to room temperature. The presence of local yielding, and the influence of coupling terms on the local stress magnitudes are examined. Extension of the results to laminated plates is presented in Part II (Dvorak, G.J., Chen, T. & Teply, J., Composites Science and Technology, 43 (1992) 359–368, this issue).     

Unrestricted Hartree-Fock approximation of the extended emery model

We use an unrestricted self-consistent Hartree-Fock approach to calculate the nature of doping states in the three-band Hubbard model. It turns out that for physically relevant parameter values one hole is localized within a small spin-polarized region where five Cu spins are aligned in the same direction. The spin polarization and binding energy between these spinpolaronic states are investigated as a function of different parameters including a Holstein-type electron-phonon coupling on the Cu sites. At higher doping concentration we observe the occurrence of afmon states where the holes are localized in a ring-shaped area. Inside this ring the antiferromagnetic order parameter has inverse sign with respect to the residual antifer-romagnetically ordered plane.     

Modelling diffusion-controlled kinetics in equilibrium and driven alloys

Summary We introduce a very simple model for thermally activated atomic migration on a lattice, and several techniques to handle it. For thermodynamical systems, the model can be used for simulating decomposition paths taking into account relevant metallurgical features. Close to equilibrium, the simplest mean-field approximation of the model can be linearized with respect to the departure from equilibrium; one then gets a microscopic interpretation of classical phenomenological coefficients, such as mobility, interfacial transfer coefficient, rate constants for the coupled relaxation of concentration and order fields. Further away from the equilibrium, the nonlinearities can be taken into account in a consistent way. For driven alloys, i.e. alloys submitted to external forcing, forced atomic migration is added to the model and new features emerge: the model is supported by several experimental results some of which confirmed its predictions a posteriori.     

Sampling average approximation method for a class of stochastic Nash equilibrium problems

We consider a class of stochastic Nash equilibrium problems (SNEP). Under some mild conditions, we reformulate the SNEP as a stochastic mixed complementarity problem (SMCP). We apply the well-known sample average approximation (SAA) method to solve the SMCP. We further introduce a semismooth Newton method to solve the SAA problems. The comprehensive convergence analysis is given as well. In addition, we demonstrate the proposed approach on a stochastic Nash equilibrium model in the wholesale gas–oil markets.     

Continuous State Dynamic Programming via Nonexpansive Approximation

This paper studies fitted value iteration for continuous state numerical dynamic programming using nonexpansive function approximators. A number of approximation schemes are discussed. The main contribution is to provide error bounds for approximate optimal policies generated by the value iteration algorithm.      

Embedding multi-dimensional meshes into twisted cubes

The twisted cube is an important variant of the most popular hypercube network for parallel processing. In this paper, we consider the problem of embedding multi-dimensional meshes into twisted cubes in a systematic way. We present a recursive method for embedding a family of disjoint multi-dimensional meshes into a twisted cube with dilation 1 and expansion 1. We also prove that a single multi-dimensional mesh can be embedded into a twisted cube with dilation 2 and expansion 1. Our work extends some previously known results.     

Stochastic approximation learning for mixtures of multivariate elliptical distributions

Most of the current approaches to mixture modeling consider mixture components from a few families of probability distributions, in particular from the Gaussian family. The reasons of these preferences can be traced to their training algorithms, typically versions of the Expectation-Maximization (EM) method. The re-estimation equations needed by this method become very complex as the mixture components depart from the simplest cases. Here we propose to use a stochastic approximation method for probabilistic mixture learning. Under this method it is straightforward to train mixtures composed by a wide range of mixture components from different families. Hence, it is a flexible alternative for mixture learning. Experimental results are presented to show the probability density and missing value estimation capabilities of our proposal.     

Combining feature spaces for classification

In this paper we offer a variational Bayes approximation to the multinomial probit model for basis expansion and kernel combination. Our model is well-founded within a hierarchical Bayesian framework and is able to instructively combine available sources of information for multinomial classification. The proposed framework enables informative integration of possibly heterogeneous sources in a multitude of ways, from the simple summation of feature expansions to weighted product of kernels, and it is shown to match and in certain cases outperform the well-known ensemble learning approaches of combining individual classifiers. At the same time the approximation reduces considerably the CPU time and resources required with respect to both the ensemble learning methods and the full Markov chain Monte Carlo, Metropolis-Hastings within Gibbs solution of our model. We present our proposed framework together with extensive experimental studies on synthetic and benchmark datasets and also for the first time report a comparison between summation and product of individual kernels as possible different methods for constructing the composite kernel matrix.