Abstract: Logging-While-Drilling (LWD) devices are often used for geosteering applications. They interpret (invert) measurements in real time to determine the well trajectory. To perform the inversion, we require a forward solver with high performance since: (a) we often need to invert for thousands of logging positions in real time, and (b) we need to solve a considerable number of forward problems. In these applications, it is a common practice to approximate the domain with a sequence of 1D models. In a 1D model, the material properties vary only along one direction (z-direction). For such 1D models, we reduce the dimensionality of the problem using a Hankel transform. We can solve the resulting system of Ordinary Differential Equations (ODEs): (a) analytically, which leads to a so-called semi-analytic method after performing a numerical inverse Hankel transform, or (b) numerically. Semi-analytic methods are used vastly due to their high performance. However, they have major limitations, namely: • By today’s knowledge, the analytical solution of the aforementioned system of ODEs is available only for piecewise-constant resistivity values. • To perform geosteering, we need to invert the measurements with respect to some inversion variables using a gradient-based inversion method. For resistivity measurements, these inversion variables are often the constant resistivity values of each layer and the bed boundary positions. However, the analytical derivatives of cross-bedded formations and the analytical derivatives of the measurements with respect to the bed boundary positions have not been published to the best of our knowledge. The main contribution of this work is to overcome the above limitations by using an efficient multi-scale finite element method to solve the system of ODEs corresponding to each Hankel mode. To do so, we divide our computations into two parts, namely: • Computations which are independent of logging positions and consist of computing the multi-scale basis functions. Hence, we precompute them once, and we use them for all logging positions. • Computations which depend upon the logging positions. Using aforementioned method, we can: (a) consider arbitrary resistivity distributions which depend upon one direction, and (b) easily and rapidly compute the derivatives with respect to any inversion variable at almost no additional cost using an adjoint state method. Although the proposed method is slower than semi-analytic ones, it is highly efficient and more flexible when computing the derivatives. In addition, the proposed method is perfectly parallelizable with respect to Hankel modes and multi-scale basis functions.

Abstract: In this paper a novel hybridization of agent-based evolu- tionary system (EMAS) is presented. This method assumes utilization of PSO for upgrading certain agents living in the EMAS population, thus serving similar to local-search methods already used in EMAS (in memetic fashion). The gathered and presented results prove the applica- bility of this hybrid based on a selection of a number of 500 dimensional benchmark functions.

Abstract: In this paper, an agent-based data-driven model that focuses on path planning layer of origin/destination popularities and route choice is developed. This model improves on the existing mathematical modeling and pattern recognition approaches. The paths and origins/destinations are extracted from a video. The parameters are calibrated from density map generated from the video. We carried out validation on the path probabilities and densities, and showed that our model generates better results than the previous approaches. To demonstrate the usefulness of the approach, we also carried out a case study on capacity analysis of a building layout based on video data.