This work discusses the advantage of using cross-correlation analysis in a data-driven approach based on principal component analysis (PCA) and piezodiagnostics to obtain successful diagnosis of events in structural health monitoring (SHM). In this sense, the identification of noisy data and outliers, as well as the management of data cleansing stages can be facilitated through the implementation of a preprocessing stage based on cross-correlation functions. Additionally, this work evidences an improvement in damage detection when the cross-correlation is included as part of the whole damage assessment approach. The proposed methodology is validated by processing data measurements from piezoelectric devices (PZT), which are used in a piezodiagnostics approach based on PCA and baseline modeling. Thus, the influence of cross-correlation analysis used in the preprocessing stage is evaluated for damage detection by means of statistical plots and self-organizing maps. Three laboratory specimens were used as test structures in order to demonstrate the validity of the methodology: (i) a carbon steel pipe section with leak and mass damage types, (ii) an aircraft wing specimen, and (iii) a blade of a commercial aircraft turbine, where damages are specified as mass-added. As the main concluding remark, the suitability of cross-correlation features combined with a PCA-based piezodiagnostic approach in order to achieve a more robust damage assessment algorithm is verified for SHM tasks.

Summary Well-control management is nowadays frequently approached by means of mathematical optimization. However, in many practical situations the optimization algorithms used are still computationally expensive. In this paper, we present progressive optimization (PO), a simulator-nonintrusivefour-stage methodology to accelerate optimal search substantially in well-controlapplications. The first stage of PO comprises a global exploration of the search space using design of experiments (DOEs). Thereafter, in the second stage, a fast-to-evaluate proxy model is constructed with the points considered in the experimental design. This proxy is based on generalized barycentric coordinates (GBCs), a generalization of the concept of barycentric coordinates used within a triangle. GBCs can be especially suited to problems in which nonlinearities are not strong, as is the case often for well-control optimization. This fact is supported by the good performance in these types of optimization problems of techniques that rely strongly on linearity assumptions, such as trajectory piecewise linearization, a procedure that is not always applicable due to its simulator-intrusive nature. In the third stage, the precision of the proxy model is iteratively improved and the enhanced surrogate model is reoptimized by means of manifold mapping (MM), a method that combines models with different levels of accuracy. MM has solid theoretical foundations and leads to efficient optimization schemes in multiple engineering disciplines. The final and fourth stage aims at additional improvement, resorting to direct optimization of the best solution from the previous stages. Nonlinear (operational) constraints are handled in PO with the filter method. The optimal search may be finalized earlier than at the fourth stage whenever the solution obtained is of satisfactory quality. PO is tested on two waterflooding problems built upon a synthetic model previously studied in well-control optimization literature. In these problems, which have 120 and 40 well controls and include nonlinear constraints, we observe for PO reductions in computational cost, for solutions of comparable quality, of approximately 30% and 50% with respect to Hooke-Jeeves direct search (HJDS), which, in turn, outperforms particle swarm optimization (PSO). HJDS and PSO are simulator-nonintrusive algorithms that usually perform well in optimization for oilfield operations. The novel concepts of GBC and MM within the framework of the PO paradigm can be extremely helpful for practitioners to efficiently deal with optimized well-control management. Savings of 50% in computing cost may be translated in practice into days of computations for just a single field and optimization run.

This paper describes the results of applying a strategy of nonlinear model predictive control (NMPC) for closed-loop optimization of the water flooding process simulation on a reversed pattern of 5 wells in the field YARIGUÍ -CANTAGALLO operated by Ecopetrol SA in Colombia.Field modeling and predictions are made through the use of a commercial reservoir simulator. The solution of the optimization problem of nonlinear control loop is determined using an approach that uses the reservoir model as a black-box by looking for patterns and sequential algorithms.To manage the information of operational variables required by the optimization strategy, an interface was developed by managing online the output files of the commercial simulator. NMPC algorithm stability is achieved by finding sub-optimal solutions of the optimization problem.The robustness and performance of the NMPC strategy is illustrated by its implementation to optimize the water flooding process simulation represented by a model with 814.226 cells of which 352.034 are active, 199 layers, 38 failures, 26 areas of balance and heterogeneous distributions of permeability and porosity. Thus, by appropriate adjustment of the water flow rates, it is proposed to increase the production of oil for the case study. Finally the main findings, conclusions and recommendations from this study are submitted.

Wells equipped with Electric Submersible Pumps (ESPs) are ubiquitous. Methodologies for either long-term (on the order of years) or short-term (months or weeks) well control are common in literature and practice. However, techniques for long-term management and optimization very often ignore devices such as ESPs in their modeling assumptions and procedures for short-term control that incorporate ESP analysis introduce important simplifications regarding fluid flow in the reservoir and well interaction. Lack of reconciliation between these long-term and short-term approaches frequently yields in practice undesired well responses and, in turn, inefficient field production. In this paper we introduce a methodology that translates controls usually considered in long-term optimization and not directly implementable in ESPs (namely, well bottomhole pressure and rate) into others that are adjustable for these devices (pump frequency and valve aperture). Fluid-flow simulation is used for reservoir modeling and ESP analysis includes viscosity correction and coupling with fluid-flow equations. The methodology allows calibration of the underlying reservoir models so that long-term and short-term reconciliation is possible. Well-control man¬ agement relies on joint optimization of drilling location and control. Calibration is formulated as an optimization problem where the discrepancy between measurements and model output is minimized through changes in reservoir parameters. The methodology is illustrated by means of a field-development and control problem constructed upon a synthetic reservoir model that has been matched with respect to 2,100 days of production. In this problem the drilling location of a new producer with an ESP and the corresponding controls for the next 1,080 days, a sequence of bottomhole pressure (BHP) values, are optimized jointly. The BHP values are then translated into ESP controls, i.e., pump frequency and valve aperture, for the following 60 days as a sequence of eight intervals, each with duration of around one week. When the ESP controls associated with the first of these intervals are simulated for the true model, the oil (water) rate at the new producer is only around 14% (40%) of the rate predicted by the matched model. Calibration of the matched model in the vicinity of the new producer results in equal rates for the true and calibrated models for the first interval. The discrepancy is relatively small for the second interval and is expected to decrease further if the (local) calibration process is repeated in future intervals. Current practice for well-control management rarely integrates joint modeling of wells systems such as ESPs and of reservoir fluid-flow dynamics. This may originate unwanted inefficiencies and losses in profitability. The comprehensive methodology for well-control management and optimization presented here aims at alleviating these modeling issues and consequently at improving production. Although described only for ESPs, the methodology is general and can be extended to other production systems.

This chapter presents an expert monitoring algorithm approach to detect, locate and quantify stiffness variations in structures. The algorithm is based on pattern recognition and artificial intelligence techniques that emulate knowledge based on human reasoning. The expert system (ES) uses time-frequency information about dynamics of structure, which is processed by using discrete wavelet transform (DWT), self-organizing maps (SOM), case-based reasoning (CBR) and principal component analysis (PCA). In addition, two applications are considered in order to evaluate the effectiveness of vibration analysis methodology and CBR in damage detection. The first application (Camacho 2010) uses the environmental excitation to detect and quantify damage in a Mechanical UBC ASCE Benchmark. The second one (Sandoval 2010) uses a predesigned signal to detect geometric damages on a gas pipeline. In both cases, a finite element model (FEM) is used to simulate different damages scenarios, which correspond to stiffness variations in different location.