Proposed Hybrid Methodology Research Articles

Multi-Objective Parameter Optimization of Fiber Laser Welding Considering Energy Consumption and Bead Geometry

Exploiting energy-saving potentials while obtaining ideal processing results is highly sought for in fiber laser welding (FLW) applications. However, little is known about the correlation between energy consumption and bead geometry of FLW, which is largely determined by processing parameters. In this study, to find the optimal processing parameters of FLW, an ensemble of variable neighborhood search–based gene expression programming (VNS-GEP) and black-box metamodels (EGBM) is presented, combined with non-dominated sorting genetic algorithm (NSGA-II) to form the proposed optimization methodology. Firstly, the optimal weight coefficients of VNS-GEP and two black-box metamodels (Kriging and SVR) are determined considering the leave-one-out generalized mean square error. Then, the EGBM is used to establish the relationship between processing parameters (laser power, welding speed and defocus distance) and response results (energy consumption and bead geometry). Additional experiments are then performed to validate the accuracy of EGBM. Analysis of variance (ANOVA) is carried out to study the main effects of the processing parameters on response results. After that, NSGA-II is employed based on EGBM to approximate the Pareto front of processing parameters with minimal total energy consumption (TEC) and maximal depth-to-width ratio (DWR). Finally, experimental validations show that the obtained solutions can achieve ideal DWR and significant reductions in TEC. In conclusion, the proposed hybrid methodology, EGBM-NSGA-II, can facilitate obtaining optimal welding processing parameters and provide a reliable empirical basis for low-energy laser manufacturing. Note to Practitioners—This study considers a practical problem of energy-aware processing parameter optimization encountered in fiber laser welding applications. Besides, the optimization methodology can be generalized to other laser processes, optimization objectives and processing conditions. This study presents an ensemble of improved gene expression programming and black-box metamodels (EGBM). An EGBM-based optimization methodology is proposed to find the optimal laser welding parameters. The ensemble combines the strengths of different metamodels, reduces the risk of using an incorrect metamodel. Note that the engineers profit from the reduced trials and time to obtain the optimal parameters for ideal results and energy-saving.

Multi-objective optimization for solving mobile robot path planning problem employing hybridization of algorithms

Mobile Robot Path Planning problem (MRPPP) is the most prominent research area employed in different real-time environments. In this article, a hybrid strategy is employed and validated with different environments for the path planning of mobile robots. The Artificial Potential Field (APF) algorithm and Multi-objective Genetic Algorithm (MOGA) are hybridized to solve an MRPPP. In the proposed hybrid methodology, the implementation is accomplished in three phases. To generate the initial population, an Artificial Potential Field (APF) algorithm is applied to discover all viable routes in an environment between the start and destination points. The collision-free paths are generated by determining the artificial force produced by obstacles and the target. The population-based evolutionary algorithm is employed to derive an optimal solution path from the initial population consisting of candidate paths. The selection operator is used to select the qualified subpopulation to the next generation to converge into the optimal solution. The genetic operators such as two-point crossover and mid-point mutation are exercised which are specific to the path planning problem.

Enhanced energy harvesting of wave energy converters in site-specific wave climates: A hybrid approach by geometric shape optimization and power take-off control

Wave energy converters are devices that capture the energy carried by ocean waves and convert it into electricity, and their performance is significantly influenced by the geometric shapes of harvesters and the control strategies of power take-off systems. This paper presents a hybrid methodology to improve conversion efficiency in realistic sea states measured from an intended deployment site, taking into account the full coupling of ocean waves, geometric shapes of wave energy harvesters, and control strategies of power take-off systems. The proposed hybrid approach employs a control-strategy-informed optimization routine to select optimal/suboptimal geometries at the design stage, then employs different control strategies for the power take-off system at the operation stage. In numerical case studies, the proposed hybrid methodology is applied to a direct-driven point-absorber type wave energy converter, where its geometric shape is optimized, informing an optimal linear passive control strategy at the design phase. Simulation results indicate that the concave shape has the best energy harvesting performance. A high capture width ratio of 428% is achieved for the found optimum concave shape that employs an optimal linear active control strategy, in a sea state with a significant wave height of 1.6 m and an average period of 5.5 s.

Hosting Capacity Calculation Deploying a Hybrid Methodology: A Case Study Concerning the Intermittent Nature of Photovoltaic Distributed Generation and the Variable Nature of Energy Consumption in a Medium Voltage Distribution Network

The main methods employed for Hosting Capacity (HC) calculations are basically classified into deterministic, stochastic and time series. In this scenario, the authors herein propose a hybrid methodology, which shows efficiency and ease of implementation. Besides the method presented, it is also calculated a hosting capacity of a real feeder which was modeled and analyzed taking into consideration variations in load and power injected by distributed generation sources. The proposed hybrid method deploys just one time series with the feeder power demand data, which are easily obtained from the feeder’s origin substation. Low voltage loads were modeled by the ratio between their maximum demands and the feeder maximum demand, making easier to start up the grid model implementation. Hence, the advantages of the proposed methodology can be summarized in: (a) easy to obtain the input parameters; (b) agility in implementing the study; (c) higher processing speed and (d) results consistent with the time series method. Finally, in view of the advantages and obtained results, the proposed hybrid methodology shows itself as a promising and attractive tool for the studies of hosting capacity by the utilities.

An effective hybrid approach based on arithmetic optimization algorithm and sine cosine algorithm for integrating battery energy storage system into distribution networks

Battery energy storage (BES) system gained more attention during integration in distribution system due to rapid development in its technology. Therefore, the best allocations and sizes of BES in radial distribution system (RDS) are determined by a recent proposed hybrid methodology combining the arithmetic optimization algorithm and sine cosine approach (AOA-SCA). The proposed approach incorporates sine cosine approach in the arithmetic algorithm to improve the exploitation phase of AOA. In this work, BES is considered in discharging state when the load demand is high and in charging state as long as the demand is low, this is managed via the proposed approach to improve the reliability of RDS. The proposed approach is applied on two different networks, modified 30-bus and IEEE 69-bus systems. The target considered in this work is minimizing the whole active power loss of the network. Other algorithms of AOA, SCA, COOT, dynamic programming (DP), and gray wolf optimizer (GWO) are implemented and compared with the proposed method. The simulation results proved the superiority of the presented AOA-SCA in identifying the best allocations and sizes of BESs in RDS. The network loss is minimized via installing one, five, ten, and sixteen BESs in IEEE 69-bus by 1.43%, 6.71%, 11.32% and 16.45%, and for modified 30-bus system by 0.42%, 2.02%, 3.33% and 4.26%, respectively.

Development of novel hybrid models for the prediction of COVID-19 in Kuwait

The first case of COVID-19 in Kuwait was reported on February 24, 2020. There is a need to develop a prediction model for estimating this epidemic size. In this study, we aimed to develop and compare several prediction models using real-time data of COVID-19 from February 24 to June 12, 2020. We modeled the uncertainty and non-stationary real-time data of COVID-19 cases using a multilayer model with different decomposition techniques. We applied our proposed hybrid methodology to predict COVID-19 cases in Kuwait. We further evaluated the performance of the novel hybrid model with others using mean relative error, mean absolute error, and mean square error. We found that our proposed hybrid approach performed better than others for predicting COVID-19 cases.

VIKOR - P-graph Method for Optimal Synthesis of Philippine Agricultural Waste-Based Sustainable Integrated Biorefinery

A sustainable integrated biorefinery (SIBR) is a biomass processing facility that converts agricultural wastes or residues into a wide range of valuable products where both economic sustainability and environmental sustainability are optimized. This work proposed a hybrid method that incorporates Vlsekriterijumska Optimizacija I Kompromisno Resenje (VIKOR), a widely used multi-criteria decision-making (MCDM) tool with the P-graph (process-graph) framework. The proposed VIKOR - P-graph model can (i) generate a set of combinatorically feasible solutions, and (ii) rank the solution sets using VIKOR method, simultaneously in the same P-graph model. In other words, all the drawbacks attributed to the sequential optimization methods can, therefore, be avoided. To demonstrate the effectiveness of the proposed hybrid methodology, a case study in the Philippines is presented in this paper. The hybrid P-graph model generated a total of 7 feasible solutions with different configuration for the SIBR based on the overall profit (i.e., economic goal) and carbon emissions (i.e., environmental goal). Results show that the best compromised solution can be obtained when majority (70 %) of the rice husk is used to produce bioethanol, where the required power is supplied by combusting the remaining rice husk and by importing external power. It offers an hourly profit of 161 $/h with a lower (~38.6 %) carbon emissions as compared to the most profitable option (3.504 tCO2-eq/h). This research is essentially a guide for policymakers to make informed decisions that can maximise the benefits of SIBR on a national scale.

Hybrid Open/Closed-Loop Attitude Control Method for Imaging Satellites

AbstractIn this work, we focus on the attitude control problem of an imaging satellite required to acquire and accurately track a target area on the Earth’s surface. The proposed hybrid methodology...

A comparison of domain integral and multi-parameter methods for the strain intensity factor estimation of rigid line inclusion using digital image correlation

A comparative study is carried out between two methods to estimate the strain intensity factor of a rigid line inclusion using the displacement and strain fields from digital image correlation (DIC). The two methods are (i) a proposed hybrid domain integral method and (ii) the linear over-deterministic least-squares technique involving multi-parameter displacement field equations (multi-parameter method). The multi-parameter method uses the displacement data obtained from the DIC experiments to find the unknowns in the multi-parameter equation. The proposed hybrid methodology uses a domain integral method to calculate the strain intensity factor using the full-field displacement and strain data obtained from DIC. The strain intensity factor estimated using the proposed domain integral method, and the multi-parameter method agrees with the analytical estimate. The influence of the size of the annular region and the mesh sensitivity is also reported. The sensitivity to the variance and bias errors in measurement for both the methods is investigated. To investigate the effect of variance error, a Gaussian noise is introduced over the reference and deformed images before post-processing, and its influence on the strain intensity factor estimate is discussed. The effect of random perturbations of displacements and strains on the strain intensity factor estimates is also investigated. The domain integral method is less sensitive to the variance error when compared to the multi-parameter solution approach. Further, the bias errors due to uncorrected lens distortions, over-smoothing of data, and DIC algorithm adopted are introduced into the measurements, and its effect on both the methods is investigated. The multi-parameter method is less sensitive to bias errors when compared to the domain integral method. Therefore, it is recommended that the domain integral method may be used for the experimental determination of strain intensity factor in measurements with variance errors and the multi-parameter method for measurements with bias errors.

A hybrid ensemble-filter wrapper feature selection approach for medical data classification

Background and objectiveMedical data plays a decisive role in disease diagnosis. The classification accuracy of high-dimensional datasets is often diminished by several redundant and irrelevant features. In this context, feature selection becomes an indispensable process. Feature selection primarily intends to identify a feature subspace which retains the classification accuracy while reducing the high computational cost of learning model as well as eliminating noise. The suitability of an appropriate feature selection approach heavily depends upon the capability of that approach to match the problem framework and to discover the intrinsic patterns within the data. The prime objective of this paper is to develop an ensemble-filter based hybrid feature selection model for disease detection. MethodsIn this paper, a four-step hybrid ensemble feature selection algorithm has been introduced. Firstly, the dataset is partitioned using the cross-validation procedure. Secondly, in the filter step, various filter methods based on weighted scores were ensembled to generate a ranking of features, and thirdly sequential forward selection algorithm is utilized as a wrapper technique to obtain an optimal subset of features. Finally, the resulting optimal subset is processed for subsequent classification tasks. ResultsExperiments have been performed on twenty benchmark medical datasets with different dimensionalities. The performance of the proposed hybrid approach is compared with fourteen state-of-the-art feature selection algorithms on four benchmark classifiers namely, Naïve Bayes, Support Vector Machine with Radial Basis Function, Random Forest and k-Nearest Neighbor. The empirical results clearly demonstrate the superiority of the proposed hybrid methodology over the competing methods with respect to accuracy, sensitivity, specificity, f1-score, area under curve evaluation measures and number of selected features. The statistical analysis of the obtained results shows the outperformance and the competitiveness of the proposed hybrid method with respect to various state-of-the-art algorithms. ConclusionsThis study concludes that the proposed hybrid approach proves to be more effective and reliable feature selection technique in selecting highly discriminative features. The framework can be utilized as a promising tool by both clinicians and researchers in enhancing the classification performance of medical datasets.

A novel hybrid grey wolf optimization algorithm and artificial neural network technique for stochastic unit commitment problem with uncertainty of wind power

The unit commitment (UC) is highly complex to solve the increasing integrations of wind farm due to intermittent wind power fluctuation in nature. This paper presents a hybrid methodology to solve the stochastic unit commitment (SUC) problem depending on binary mixed integer generator combination with renewable energy sources (RESs). In this combination, ON/OFF tasks of the generators are likewise included to satisfy the load requirement as for the system constraints. The proposed hybrid methodology is the consolidation of grey wolf optimization algorithm (GWOA) and artificial neural network (ANN), hence it is called the hybrid GWOANN (HGWOANN) technique. Here, the GWOA algorithm is used to optimizing the best combination of thermal generators depending on uncertain wind power, minimum operating cost and system constraints – that is, thermal generators limits, start-up cost, ramp-up time, ramp-down time, etc. ANN is utilized to capture the uncertain wind power events, therefore the system ensures maximal application of wind power. The combination of HGWOANN technique guarantees the prominent use of sustainable power sources to diminish the thermal generators unit operating cost. The proposed technique is implemented in MATLAB/Simulink site and the efficiency is assessed with different existing methods. The comparative analysis demonstrates that the proposed HGWOANN approach is proficient to solve unit commitment problems and wind integration. Here, the HGWOANN method is compared with existing techniques such as PSO, BPSO, IGSA to assess the overall performance using various metrics viz. RMSE, MAPE, MBE under 50 and 100 count of trials. In the proposed approach, the range of RMSE achieves 9.26%, MAPE achieves 0.95%, MBE achieves 1% in 50 count of trials. Moreover, in 100 count of trials, the range of RMSE achieves 7.38%, MAPE achieves 1.91%, MBE achieves 2.87%.

Integrating Domain Knowledge in Data-Driven Earth Observation With Process Convolutions

The modelling of Earth observation data is a challenging problem, typically approached by either purely mechanistic or purely data-driven methods. Mechanistic models encode the domain knowledge and physical rules governing the system. Such models, however, need the correct specification of all interactions between variables in the problem and the appropriate parameterization is a challenge in itself. On the other hand, machine learning approaches are flexible data-driven tools, able to approximate arbitrarily complex functions, but lack interpretability and struggle when data is scarce or in extrapolation regimes. In this paper, we argue that hybrid learning schemes that combine both approaches can address all these issues efficiently. We introduce Gaussian process (GP) convolution models for hybrid modelling in Earth observation (EO) problems. We specifically propose the use of a class of GP convolution models called latent force models (LFMs) for EO time series modelling, analysis and understanding. LFMs are hybrid models that incorporate physical knowledge encoded in differential equations into a multioutput GP model. LFMs can transfer information across time-series, cope with missing observations, infer explicit latent functions forcing the system, and learn parameterizations which are very helpful for system analysis and interpretability. We consider time series of soil moisture from active (ASCAT) and passive (SMOS, AMSR2) microwave satellites. We show how assuming a first order differential equation as governing equation, the model automatically estimates the e-folding time or decay rate related to soil moisture persistence and discovers latent forces related to precipitation. The proposed hybrid methodology reconciles the two main approaches in remote sensing parameter estimation by blending statistical learning and mechanistic modeling.

Routing in a Wireless Sensor Network using a Hybrid Algorithm to Improve the Lifetime of the Nodes

A reliable and secure routing protocol for Wireless Sensor Networks should be easy to maintain, reliable and cost-efficient. In this paper, we propose a hybrid routing algorithm using Ant Colony Optimization and Minimum Hop Count scheme. The proposed hybrid methodology provides an optimal routing path that ensures balanced and minimal energy consumption. The hybrid algorithm is unique in maintaining network topology, balancing network load and searching for the optimal route. The proposed algorithm with the WSN model is implemented in C++ and the simulation output proves our algorithm to outperform other similar routing algorithms.The proposed work indicates animprovement in network lifetime, success rate in finding the best solution and rate of convergence. The existing techniques are reviewed and their strengths and weaknesses are diagnosed and compared with our proposed hybrid methodology that integrates the strengths of both the algorithms.

WITHDRAWN: A hybrid approach for classification of infectious diseases

Infectious diseases like typhoides, dengues, etc. are caused by water and environmental changes. It is important to provide medications for such infection patients. Conventional approaches for the correct diagnosis of patients are not acceptable. To make an appropriate diagnosis it is important to accurately identify patient status. Although advancement in reliable diagnostics of technology remains an essential mission. This thesis suggested a classification approach for certain infectious diseases. 173 Typhoid fever patients and 119 Dengue Fever patients were collected clinically. Crafted a hybrid method using voting set technique that was constructed as base classification in machine learning techniques. The diagnostic results can be classified according to the classificator's majority of votes. Data set tests were carried out and the results were contrasted with traditional computer teaching methods, including the Group Baïve Bayesian, K-Nearest Neighbor and the Decision Tree. Analysis indicates that, in the case of typhoid and dengue-fever infected patients, the suggested hybrid solution has a 97.2% precision. Focusing on infectious diseases, several traditional approaches were analysed and the proposed hybrid methodology focused on an ensemble voting strategy that showed the highest efficiency in classifying infectious diseases.

LoRD-Net: Unfolded Deep Detection Network With Low-Resolution Receivers

The need to recover high-dimensional signals from their noisy low-resolution quantized measurements is widely encountered in communications and sensing. In this paper, we focus on the extreme case of one-bit quantizers, and propose a deep detector entitled LoRD-Net for recovering information symbols from one-bit measurements. Our method is a model-aware data-driven architecture based on deep unfolding of first-order optimization iterations. LoRD-Net has a task-based architecture dedicated to recovering the underlying signal of interest from the one-bit noisy measurements without requiring prior knowledge of the channel matrix through which the one-bit measurements are obtained. The proposed deep detector has much fewer parameters compared to black-box deep networks due to the incorporation of domain-knowledge in the design of its architecture, allowing it to operate in a data-driven fashion while benefiting from the flexibility, versatility, and reliability of model-based optimization methods. LoRD-Net operates in a blind fashion, which requires addressing both the non-linear nature of the data-acquisition system as well as identifying a proper optimization objective for signal recovery. Accordingly, we propose a two-stage training method for LoRD-Net, in which the first stage is dedicated to identifying the proper form of the optimization process to unfold, while the latter trains the resulting model in an end-to-end manner. We numerically evaluate the proposed receiver architecture for one-bit signal recovery in wireless communications and demonstrate that the proposed hybrid methodology outperforms both data-driven and model-based state-of-the-art methods, while utilizing small datasets, on the order of merely $\sim 500$ samples, for training.

A continuous data driven translational model to evaluate effectiveness of population-level health interventions: case study, smoking ban in public places on hospital admissions for acute coronary events

BackgroundAn important task in developing accurate public health intervention evaluation methods based on historical interrupted time series (ITS) records is to determine the exact lag time between pre- and post-intervention. We propose a novel continuous transitional data-driven hybrid methodology using a non-linear approach based on a combination of stochastic and artificial intelligence methods that facilitate the evaluation of ITS data without knowledge of lag time. Understanding the influence of implemented intervention on outcome(s) is imperative for decision makers in order to manage health systems accurately and in a timely manner.MethodsTo validate a developed hybrid model, we used, as an example, a published dataset based on a real health problem on the effects of the Italian smoking ban in public spaces on hospital admissions for acute coronary events. We employed a continuous methodology based on data preprocessing to identify linear and nonlinear components in which autoregressive moving average and generalized structure group method of data handling were combined to model stochastic and nonlinear components of ITS. We analyzed the rate of admission for acute coronary events from January 2002 to November 2006 using this new data-driven hybrid methodology that allowed for long-term outcome prediction.ResultsOur results showed the Pearson correlation coefficient of the proposed combined transitional data-driven model exhibited an average of 17.74% enhancement from the single stochastic model and 2.05% from the nonlinear model. In addition, data demonstrated that the developed model improved the mean absolute percentage error and correlation coefficient values for which 2.77% and 0.89 were found compared to 4.02% and 0.76, respectively. Importantly, this model does not use any predefined lag time between pre- and post-intervention.ConclusionsMost of the previous studies employed the linear regression and considered a lag time to interpret the impact of intervention on public health outcome. The proposed hybrid methodology improved ITS prediction from conventional methods and could be used as a reliable alternative in public health intervention evaluation.

Hybrid Methodology for Efficient on the Fly (Re)Parametrization of Proton Exchange Membrane Fuel Cells Electrochemical Model for Diagnostics and Control Applications

Prevention and diminution in effects of degradation phenomena under highly dynamic operation of proton exchange membrane fuel cells (PEMFC), while retaining sufficiently high efficiency and performance of the system, are considered significant obstacles toward their wider use in transport applications. Overcoming these obstacles calls for precise on-line monitoring and control tools such as coupled observers, which enable combined performance and service life optimizations. Models used in these applications should feature low computational effort, good extrapolation capabilities and should be easy to parametrize. The first requirement, which is obvious for on-line applications, in frequently in contradiction to the last two, which are heavily intertwined. Good extrapolation capabilities of the model significantly reduce the size of experimental data sets needed for successful parametrization, thus enabling the parametrization on small data sets, while retaining higher accuracy outside of trained variation space. This rationale motivates the use of the computationally-fast reduced-dimensionality electrochemical models e.g. [1, 2], featuring a more profound mechanistic basis; thus, exhibiting better prediction capability of the model.To present significant progress in the aforementioned area, this contribution presents framework of computationally-fast electrochemical models that can be parametrized whether on experimental data from polarization curves or on data obtained by electrochemical impedance spectroscopy (EIS). This hybrid methodology consists of interchangeable thermodynamically consistent reduced-dimensionality electrochemical model for PEMFC [2] and newly developed analytical physical model for impedance spectra with the same set of calibration parameters. The latter is developed by applying electrochemical ansatzes for the cathode and anode reactions derived in [2] on the proposed modelling approach used for calculation of the cathode catalyst layer impedance in the inspiring work of Kulikovsky [3]. Profound modelling basis using the same set of calibration parameters enable enhanced identification of individual calibration parameters that are otherwise harder to be uniquely determined [4] on one hand, and enable parametrization of the model from either of the two separate measurements in time and frequency domains on the other.Enhanced identifiability of calibration parameters using proposed hybrid methodology is especially pronounced in the case of parameters describing anodic and cathodic reaction rates and their activation energies, which are extremely hard to be uniquely determined based on the experimental data consisting of polarization curves only, due to coupling of their information with other calibration parameters. Additionally, proposed hybrid methodology also enables a significant reduction in the measurement time in the case of experimental data obtained by EIS, allowing the low frequency regions, where the transport phenomena could be otherwise characterized, to be omitted due to already uniquely determined calibration parameters from polarization curve measurements. Furthermore, to reduce the necessary amount of experiments for successful parametrization of the model even further, the optimal design of experiments (DoE) is used, based on D-optimal criterion applied onto Fisher information matrix.To assess developed hybrid methodology calibration parameters, obtained by means of minimizing penalty function value with differential evolution algorithm on EIS data, are used for validation against polarization curves measured at the same RuL (remaining useful life). Furthermore, to confirm applicability of the methodology, the same procedure is also used for calibration on the measured polarization curves and validation on the EIS experimental data. Small discrepancy between the values of the root-mean-square errors show plausibility of the proposed hybrid methodology. On the other hand, Fisher information analysis is performed to assess increment in the information about individual calibration parameter when each experimental point from EIS data set is added to experimental space used for parametrization.Results of this joint analysis offer unique insight into the information that data set obtained with the EIS measurements possess about individual calibration parameter in the individual frequency interval and intuitively confirm the results obtained by application of the optimal DoE for a synthetic space of EIS measurements. The latter extremely reduces experimental workload needed for successful calibration of the model in comparison with the full factorial. Innovative hybrid methodology of the measurements in time and frequency domains can thus be used for achieving reduced costs and efforts in initial parametrization and during operation in re-parametrization procedures of the modelling framework for virtual observers.

Hybrid Methodology for Efficient on the Fly (Re)Parametrization of Proton Exchange Membrane Fuel Cells Electrochemical Model for Diagnostics and Control Applications

Successful parametrization and re-parametrization of the models used in PEMFC observer applications is instrumental to assure ideal control of the system. To increase ease of parametrisation and accuracy of parameter determination, this paper presents framework of twin analytical ansatzes for modelling PEMFC response in time and frequency domains sharing same set of calibration parameters. Owing to thermodynamically consistent modelling basis of the twin models, which are based on electrochemical model with state-of-the-art extrapolation capabilities and replication of the experimental data with one set of calibration parameters, they are valid in all current density regions. Furthermore, unique sharing of the calibration parameters enables unprecedented enrichment of the dataset that can be used to determine values of model’s calibration parameters with higher certainty or enhances identification of individual calibration parameters that are otherwise harder to be uniquely determined. Additionally, proposed hybrid methodology also enables a significant reduction in the measurement time and enable re-parametrization on the fly.

A hybrid intelligent model for reservoir production and associated dynamic risks

This research presents a hybrid model to predict oil production and to provide a dynamic risk profile of the production system. The introduced predictive approach combines a multilayer perceptron (MLP)-artificial neural network (ANN) model with a hybrid connectionist strategy (BN-DBN), which comprises a Bayesian network (BN) model and a dynamic Bayesian network (DBN) model. The proposed hybrid methodology (MLP-BN-DBN) is designed to find the correlations between the input and output data to forecast the desired oil production rate. The MLP model captures the variabilities in the fluid and rock properties, model's uncertainties, and the effects of pressure maintenance on the production process. The BN model uses the 3σ mathematical rule to promptly signal the arrival of any production rate change and captures the pressure maintenance impact using the early warning source indexes. The DBN model provides a dynamic risk profile of the production system using the observed evidence and reservoir production hyperbolic decline concept.The proposed methodology offers the field operators better opportunity to obtain real time estimate of the likelihood of any impending production loss at any time during production operations. The model exhibits a high capability of oil production prediction with the minimum, average, and maximum percentage errors of 0.01%, 6.57%, and 15.28%, respectively. The developed hybrid model serves as a risk monitoring system. The model is cost-effective and eases the computational burden of history matching processes and bridges the gaps in the existing systems for oilfield development dynamic risk forecast and production predictions. Hence, the proposed methodology offers a multipurpose tool for dynamic risk assessment and for proper reservoir production optimization.

A hybrid methodology to map informal settlements in Durban, South Africa

Currently, given the rapid growth and implementation of inadequate planning policies in African cities, the gap between formal and informal urbanisation is becoming increasingly evident. Spontaneous and unplanned settlements are home to more than half the urban population in Africa, where lost and forgotten spaces within cities are claimed by the urban poor as their ‘right to the city’. This study proposes an innovative research methodology of spatial mapping to analyse informal settlements, thus giving them a new ‘lived’ identity. The research has two key objectives: understanding how lived space can be effectively captured and interpreted and using spatial mapping as a key methodological approach to reinterpreting and adapting spontaneous settlements to existing challenges, hazards and risks. The proposed hybrid methodology, which combines drone photography and qualitative research methods (e.g. interviews, collaborative mapping), was tested in two pilot studies in the Durban metropolitan area, South Africa, to reveal the real attributes and vulnerabilities of such human settlements and to understand how they can be reorganised and adapted. The outcomes of this study allow for a more accurate (lived four-dimensional) mapping of scenarios affecting cities today.