Figure 13 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
The data used here are digitized from Yungul (1950) algong A–M profile, it is called weiss anomaly. doi:10.1371/journal.pone.0051199.g013
Source publication
Differential evolution (DE) is a population based evolutionary algorithm widely used for solving multidimensional global optimization problems over continuous spaces, and has been successfully used to solve several kinds of problems. In this paper, differential evolution is used for quantitative interpretation of self-potential data in geophysics....
Citations
... where i ≠ r 1 ≠ r 2 ≠ r 3 , r 1 , r 2 , r 3 are random integers in the interval [1,NP], F is the scaling factor, g denotes the number of evolutionary generations, and X i (g) denotes the ith individual in the gth generation population (Li and Yin 2012). By mutation, the gth generation population produces a new intermediate population V i (g + 1). ...
The cement industry is one of the main sources of NOx emissions, and automated denitration systems enable precise control of NOx emission concentration. With non-linearity, time delay and strong coupling data in cement production process, making it difficult to maintain stable control of the denitration system. However, excessive pursuit of denitration efficiency is often prone to large ammonia escape, causing environmental pollution. A multi-objective prediction model combining time series and a bi-directional long short-term memory network (MT-BiLSTM) is proposed to solve the data problem of the denitration system and achieve simultaneous prediction of NOx emission concentration and ammonia escape value. Based on this model, a model predictive control framework is proposed and a control strategy of denitration system with multi-index model predictive control (MI-MPC) is built based on neural networks. In addition, the differential evolution (DE) algorithm is used for rolling optimization to find the optimal solution and to obtain the best control variable parameters. The control method proposed has significant advantages over the traditional PID (proportional integral derivative) controller, with a 3.84% reduction in overshoot and a 3.04% reduction in regulation time. Experiments prove that the predictive control framework proposed in this paper has better stability and higher accuracy, with practical research significance.
... However, global optimizers generally require more computational resources than local optimizers. Several studies have reported on the inversion of SP anomalies using global optimizers (Gobashy & Abdelazeem, 2021), including genetic algorithm (GA) (Abdelazeem & Gobashy, 2006;Gö ktü rkler & Balkaya, 2012), particle swarm optimization (PSO) (Sweilam et al., 2007;Santos, 2010;Gö ktü rkler & Balkaya, 2012;Essa, 2019;Ekinci et al., 2020), differential evolution (DE) (Li & Yin, 2012;Balkaya, 2013), micro-differential evolution (MDE) (Sungkono, 2020), simulated annealing (SA) (Gö ktü rkler & Balkaya, 2012), very fast simulated annealing (VFSA) (Biswas, 2017;Biswas et al., 2022), genetic price algorithm (GP) (Di Maio et al., 2016), crow search algorithm (CSA) (Haryono et al., 2020), whale optimization algorithm (WOA) (Abdelazeem et al., 2019;Gobashy et al., 2020), black hole algorithm (BHA) (Sungkono, 2018), cuckoo search algorithm (CSA) (Turan-Karaog lan & Gö ktü rkler, 2021), bat optimizing algorithm (BOA) (Essa et al., 2023), and self-adaptive bare-bones teaching-learning-based optimization (SABBTLBO) (Sungkono, 2023). However, according to the No Free Lunch Theorem of optimization (Wolpert & Macready, 1997), there is no definite algorithm that can handle all types of inverse problems. ...
The self-potential method (SP) has been used extensively to reveal some model parameters
of various ore deposits. However, estimating these parameters can be challenging due to the
mathematical nature of the inversion process. To address this issue, we propose here a novel
global optimizer called the Modified Barnacles Mating Optimizer (MBMO). We improved
upon the original approach by incorporating a variable genital length strategy, a novel
barnacle offspring evolving method, and an out-of-bounds correction approach. The MBMO
has not been previously applied to geophysical anomalies. Prior to inversion of real data sets,
modal and sensitivity Analyzes were conducted using a theoretical model with multiple
sources. The Analyzes revealed that the problem is modal in nature, model parameters have
varying levels of sensitivity, and an algorithm that can well balance global exploration with
local exploitation is required to solve this problem. The MBMO was tested on theoretical SP
anomalies and four real datasets from Tu¨ rkiye, Canada, India, and Germany. Its performance
was compared to the original version under equal conditions. Uncertainty determination
studies were carried out to comprehend the reliability of the solutions obtained via
both algorithms. The findings indicated clearly that the MBMO outperformed its original
version in estimating the model parameters from SP anomalies. The modifications presented
here improved its ability to search for the global minimum effectively. In addition to geophysical
datasets, experiments with 11 challenging benchmark functions demonstrated the
advantages of MBMO in optimization problems. Theoretical and field data applications
showed that the proposed algorithm can be used effectively in model parameter estimations
from SP anomalies of ore deposits with the help of total gradient anomalies.
... Langkah selanjutnya adalah memilih sejumlah IMF sebagai data terfilter yang terbaik. Dalam proses pemilihan ini, hanya IMF1 yang diabaikan dengan beberapa pertimbangan, antara lain [17]: 1) IMF1 memiliki panjang gelombang pendek yang disebabkan oleh heterogenitas tanah, 2) Sesar Grindulu dibentuk oleh reaktivasi patahan di batuan dasar, sehingga sesar ini diperlihatkan oleh data SP dengan panjang gelombang yang panjang, 3) anomali regional dari data SP pada umumnya memiliki nilai konstan atau linier [18], [19]. ...
Self-Potential (SP) is a geophysical survey method that is relatively easy and inexpensive. Interpretation of SP data can be used for various purposes such as the detection of landslide-prone areas, exploration of various types of minerals, and identification of the parameters of a fault or crack. In this study, SP data acquisition was carried out in Tambakrejo Village, Pacitan District with a total of 102 measurement data which aims to determine the depth and dip of the Grindulu Fault. SP data acquired in the field needs to be corrected for reference, namely corrections caused by a displacement of the starting point of measurement. This data is then filtered to increase the signal-to-noise ratio (SNR) and sharpen the resulting anomalies. This filtering process is carried out using the ICEEMD (Improved Complete Ensemble Empirical Mode Decomposition) method which is a development of the EMD method. Furthermore, the SP data inversion process to obtain model parameters is carried out by utilizing the CSA (Crow Search Algorithm) method. Based on the anomaly model generated from the SP data inversion process, it can be concluded that the Grindulu Fault was identified at a distance of 803,8 meters from the starting point of measurement with depths ranging from 11,06 to 102,74 meters. Furthermore, based on distance, depth, and anomaly shape data, the dip value can be calculated. The calculation results show that the dip of the Grindulu Fault in the study area is 75.58o. Identification of the Grindulu Fault in the form of depth and dip is very important in efforts to model the fault comprehensively.
... Recently, with the use of idealized body anomaly assumption as the source, SP data inversion using GO methods has seen rapid development, including the genetic algorithm (GA) (Göktürkler & Balkaya, 2012), simulated annealing (SA) (Sharma & Biswas, 2013), differential evolution (DE) (Balkaya, 2013;Li & Yin, 2012;Sungkono, 2020a), particle swarm optimization (PSO) (Essa, 2019(Essa, , 2020Fernández-Martínez et al., 2010;Monteiro Santos, 2010), black hole algorithm (BHA) (Sungkono & Warnana, 2018), whale algorithm (WA) (Abdelazeem et al., 2019), crow search algorithm (CSA) (Haryono et al., 2020), genetic price algorithm (GPA) (Di Maio et al., 2019), flower pollination algorithm (FPA) (Sungkono, 2020b), cuckoo search algorithm (Karaoglan & Göktürkler, 2021), symbiosis optimization search (SOS) (Sungkono & Grandis, 2021), and bat optimizing algorithm (BOA) (Essa et al., 2022). Moreover, grey wolf optimization (GWO), salp swarm optimization (SSO), and artificial bee colony (ABC) have been successfully employed for singleanomaly SP data inversion (Gobashy & Abdelazeem, 2021). ...
Teaching–learning-based optimization (TLBO) is a meta-heuristic algorithm that simulates the process of teacher and student (or learner) interaction in transmitting knowledge. The algorithm is relatively simple to implement, with free-tuning parameters for balancing exploration and exploitation of the solution space. TLBO contains two phases, namely, teaching and learning. In this paper, self-adaptive Gaussian bare-bones TLBO (SABBTLBO) is proposed for improving TLBO and Gaussian bare-bones TLBO (BBTLBO) performance. In the SABBTLBO, Gaussian bare-bones and the original teaching phase in TLBO become more adaptive by a mechanism based on the learner’s rank. For the new learning phase, an adaptive scaling factor based on the rank mechanism is used to modify the neighborhood search strategy. A restarted mutation approach is also added in the learning phase. The developed SABBTLBO is compared with six state-of-the-art TLBO variant algorithms for inversion of synthetic multiple self-potential (SP) anomaly sources. The proposed SABBTLBO algorithm is also tested and compared with several algorithms applied for field SP data from different locations in the world including India, Portugal, and Indonesia, using the assumption that SP data are sourced by idealized bodies (simple geometric model or thin sheet model). The inversion of multiple SP anomaly sources using SABBTLBO is used not only for determining the best model parameters, but also their uncertainties. The latter is estimated from the equivalence region of the set of possible solutions via cost function topography evaluation. Significant results were obtained and can be associated with the geology of studied area.
... In recent decades, non-derivative natureinspired global optimization and metaheuristics have become prevalent, rather than using derivative-based local-search optimization to resolve geophysical inverse problems [56,57]. SP data have been interpreted using various global optimization algorithms, including genetic algorithms [23], neural network [28], particle swarm optimization [39,58], differential evolution [59,60], ant colony optimization [61], black-hole algorithm [62], genetic-price algorithm [27], and micro differential evolution algorithm [63]. These optimization algorithms have been well applied in the interpretation of idealized bodies such as spheres, horizontal and vertical cylinders, and thin and thick sheets for SP data. ...
Self-Potential data have been widely used in numerous applications. The interpretation of SP data from subsurface bodies is quite challenging. The advantages of geophysical inversion for interpreting non-linear geophysical problems have gained a great deal of attention over conventional interpretation. The efficiency of the present inversion approach in interpreting SP anomalies from a thin dipping layer/bed is presented in the study. The inversion approach was applied to interpret synthetic model parameters such as the self-potential of the layer (k), depth to the body top (h), location of the body (x0), dip angle (θ), and the upper and lower end of the sheet (δ1 and δ2). The interpretation of the results showed that the parameters Δh, δ1, and δ2 exhibited a wide range of results. The estimated parameter values lay within the limit of uncertainty. The inversion approach was also applied to two field datasets obtained from polymetallic deposits in Russia and Azerbaijan for mineral exploration purposes and one from a buried ancient Roman limestone construction in Halutza, Israel, for the purposes of archaeological study. The field investigation results demonstrate a good agreement with previous works of literature. The efficiency of the present approach for interpreting SP anomalies from thin layer/bed-like structures is shown in this study.
... Literatürde sıklıkla kullanılan metasezgisel yöntemler genetik algoritma-GA [4], yapay ısıl işlem-YIİ [5], parçacık sürü optimizasyonu-PSO [6], farksal evrim-FE [7][8][9] olarak örneklendirilebilir. Jeofizikte parametre kestirim çalışmalarında yaygın olarak kullanılan algoritmalara ise GA [10][11][12], PSO [13][14][15][16][17][18][19], FE [2,3,[19][20][21][22][23][24], YIİ algoritması [25][26][27][28][29] farksal arama algoritması [30,31] örnek olarak verilebilir. Parametre kestirim çalışmalarında uygulanan güncel algoritmalara örnekler ise gri kurt optimizasyonu [32] geri-izleme arama optimizasyonu [33], guguk kuşu arama algoritması [34] olarak verilebilir. ...
Metasezgisel algoritmalar jeofizik ters çözüm çalışmalarında sıklıkla kullanılır duruma gelmiştir. Türev tabanlı en iyileme yöntemlerinin aksine, iyi bir başlangıç modeline ihtiyaç duymayan arama algoritmaları parametre uzayını kapsamlı tarama özelliğine sahip olduklarından jeofizikte model parametre kestirimleri için avantaj sağlamaktadır. Sunulan çalışmada, gravite anomalilerinin ters çözümünde guguk kuşu arama algoritması kullanılmıştır. Algoritmanın kullanıcı tanımlı parametre sayısının az olması ve yapılan literatür taramasında doğadan esinlenilerek oluşturulan birçok metasezgisel yönteme göre daha iyi sonuç vermesi, parametre kestirim çalışmasında guguk kuşu algoritmasının kullanılmasını teşvik etmektedir. Gravite belirtisine ait genlik katsayısı, kaynak derinliği, belirti izdüşümü ve şekil faktörleri kestirimi yapılan model parametreleridir. Algoritmaya ait kontrol parametreleri (popülasyon sayısı ve yumurtanın yuvadan atılma olasılığı) ise gürültüsüz kuramsal veri kümesi kullanılarak parametre belirleme çalışmaları (parameter tuning) ile detaylı bir şekilde irdelenmiştir. Sonrasında kontrol parametre çiftinin doğruluğu gürültü içeren veri kümesi üzerinde test edilmiştir. Ardından, Küba’da bir kromit yatağı üzerinde ölçülen arazi verisi ve Kanada’da yer alan bir sülfit cevheri üzerinde ölçülen arazi verisi değerlendirilerek, anomalilere ait model parametreleri kestirilmiştir. Kuramsal ve arazi veri kümelerine ait model parametrelerinin güvenilirliğinin belirlenmesi için, Metropolis-Hasting algoritması kullanılarak, kestirim parametreleri istatistiksel olarak da test edilmiştir. Doğası gereği iyi bir başlangıç modeline ve model parametrelerine göre kısmi türev hesabına ihtiyaç duymayan algoritma, kullanıcı tanımlı iki parametre içermesi sayesinde parametre kestirim çalışmalarında kolaylık sağlamıştır. Yapılan belirsizlik analizleri sonucunda da algoritmanın gravite verilerinin ters çözümünde uygulanabilir bir algoritma olduğu belirlenmiştir.
... Among this, is the introduction of voxel based inversion programs which use regularization techniques to generate geologically reasonable models (Li and Oldenburg 1996. The implementation of compression techniques to improve the computational performance of the algorithms (Li and Oldenburg 2003) and in the field of self-potential inversion, many progress can be observed where tomographic methods are presented (Di Maio and Patella 1994;Patella 1997;Di Maio et al. 2013, 2016a, least squares inversion (Abdelrahman et al. 2008;Li and Yin 2012), spectral analysis (Rani et al. 2015), and global optimization approaches (Tlas and Asfahani 2008;Srivastava et al. 2014;Di Maio et al. 2016b). ...
... Application of metaheuristics in SP inversion is mainly belongs to the first and second categories, EA's (Gobashy et al. 2019;Abdelazeem and Gobashy 2006;Li and Yin 2012;Di Maio et al. 2016b;Balkaya et al. 2017) and SI's (Sweilam et al. 2007;Srivastava et al. 2014;Singh and Biswas 2016). The application of AIS in SP or geophysical inversion is not yet notices in any literature.The physics based metaheuristics are presented in SP inversion by the simulated annealing (SA), as in Sharma 2014, 2015). ...
... Three control parameters are the only requirements: number of population (N p), weighting factor (mutation constant, F) and crossover probability (Cr). The initial population is generated randomly in the initialization stage of the algorithm, then in the evolution stage population evolves from one generation to the next through mutation, crossover and selection operations until the termination criterion is satisfied (Fig. 4.3) (Li and Yin 2012;Ekinci et al. 2016). The target vectors can be defined as 2, . . . ...
Artificial intelligence and metaheuristic approaches had gained a remarkable position in last‐millennium geophysical inversion. The past two decades have witnessed the development of numerous metaheuristics in various communities that sit at the intersection of several fields including geophysics. Many inverse problems in geophysics are considered as constrained optimization, as the aim of the process is to find the best parameter estimates so as to minimize the differences between the predicted results and the observations while satisfying all known constraints or thresholds. Such optimization problems can thus be solved by efficient traditional optimization techniques (e.g. Least-squares). However, as the number of degrees of freedom is usually very large, metaheuristic algorithms such as, Whale, Grey Wolf, particle swarm, genetic, Bat, and Cuckoo Search algorithms are particularly suitable for inverse problems of that kind, because metaheuristics are very efficient for solving non-linear global optimization problems. The inversion of spontaneous potential (SP) anomalies in particular, attracted many authors because of its applicability in many fields of applied geophysics, such as mining, archaeology, dam seepage and many others. This chapter provides a complete view of metaheuristics as effective tool for parameter estimation from the SP signal. We show the main design questions and search components for selected families of metaheuristics. Not only the design aspect of metaheuristics but also their implementation including the formulation of the Objective/target function. After covering the common synthetic examples with the noise tests, many field examples will be presented to show its effectiveness and suitability for various geologic conditions and a diverse range of application domains.
... Li et al. [128] applied DE to estimate the self-potential data widely used in geophysics with better solution quality and efficiency by optimising six parameters, namely, regional coefficients, polarisation angle, distance from the origin, depth of source and electrical dipole moment. Marcˇicˇet al. [129] proposed to use DE to simultaneously identify the mechanical, electrical and magnetic subsystem parameters of a dynamic model used to represent the objective functions and technical constraints of line-start interior permanent magnet synchronous motors. ...
... Year Applications [124] 2016 Prediction [125] 2018 Prediction [126] 2018 Prediction [127] 2019 Prediction [128] 2012 Industrial control [129] 2014 Industrial control [130] 2015 Industrial control [131] 2015 Industrial control [132] 2019 Industrial control [133] 2020 Industrial control [134] 2017 Computational systems [135] 2020 Computational systems [136] 2020 Computational systems [137] 2012 Electrical and power systems [138] 2017 Electrical and power systems [139] 2017 Electrical and power systems [140] 2019 Electrical and power systems [141] 2020 Electrical and power systems [142] 2013 Feature selection [143] 2017 Feature selection [144] 2017 Feature selection [145] 2018 Feature selection [146] 2018 Feature selection [147] 2018 Feature selection [148] 2020 Feature selection [149] 2020 Feature selection [150] 2020 Feature selection [151] 2013 Image processing [152] 2017 Image processing [153] 2018 Image processing [154] 2019 Image processing [155] 2020 Image processing [156] 2017 Clustering [157] 2019 Clustering [158] 2019 Clustering [159] 2019 Clustering [160] 2020 Clustering [161] 2018 Health care [162] 2019 Health care [163] 2019 Health care [164] 2016 Path planning [165] 2019 Path planning [166] 2020 Path planning [167] 2020 Path planning [168] 2020 Path planning [169] 2020 Path planning [170] 2018 Wireless and sensor [171] 2018 Wireless and sensor [172] 2018 Wireless and sensor [173] 2020 Wireless and sensor [174] 2007 Differential equations [175] 2013 Differential equations [176] 2014 Differential equations [177] 2019 Differential equations [178] 2020 Differential equations 4.6.5. Feature selection Ghosh et al. [142] proposed a self-adaptive DE (SADE) to address the feature subset selection problem of a hyperspectral image that suffers from high computational intensiveness and redundancy issues due to the presence of large numbers of neighbouring bands. ...
Differential evolution (DE) is a popular evolutionary algorithm inspired by Darwin's theory of evolution and has been studied extensively to solve different areas of optimisation and engineering applications since its introduction by Storn in 1997. This study aims to review the massive progress of DE in the research community by analysing the 192 articles published on this subject from 1997 to 2021, particularly studies in the past five years. The methodology used to search for relevant DE papers and an overview of the original DE are firstly explained. Recent advances in the modifications proposed to enhance the effectiveness and efficiency of the original DE are reviewed by analysing the strengths and weaknesses of each published work, followed by the potential applications of these DE variants in solving different real-world engineering problems. In contrast to most existing DE review papers, additional analyses are performed in this survey by investigating the impacts of various parameter settings on given DE variants to identify their optimal values required for solving certain problem classes. The qualities of modifications incorporated into selected DE variants are also evaluated by measuring the performance gains achieved in terms of search accuracy and/or efficiency against the original DE. The additional surveys conducted in this study are anticipated to provide more insightful perspectives for both beginners and experts of DE research, enabling their better understanding about current research trends and new motivations to outline appropriate strategic planning for future development works. Ó 2021 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
... In practice, EAs are a family of nature-inspired algorithms widely used for solving complex optimization problems which can be used for assisting developers in determining the optimal parameters of the training model. The differential evolution (DE) algorithm [24][25][26][27][28] is a branch of EA that follows the general procedures of EAs. ...
... The fundamental idea behind DE is a scheme for producing trial vectors according to the manipulation of target vector and difference vector. If the trail vector yields a lower objective function than a predetermined population member, the newly generated trail vector will replace the vector and be compared in the following generation [28]. ...
... Following the architecture analysis step of detailed model design, experiments were conducted with model training in order to validate the optimal parameters of the developed TCN model based on training samples. In the experiment, differential evolution (DE) methods [24][25][26][27][28] explored these solutions to handle the hyper-parameter tuning of the TCN model for predicting wind power output in order to reach the satisfactory prediction accuracy for different weather conditions. Essentially, the DE method is a population-based stochastic search process using the distance and direction information from the current population to conduct its search. ...
Studies have demonstrated that changes in the climate affect wind power forecasting under different weather conditions. Theoretically, accurate prediction of both wind power output and weather changes using statistics-based prediction models is difficult. In practice, traditional machine learning models can perform long-term wind power forecasting with a mean absolute percentage error (MAPE) of 10% to 17%, which does not meet the engineering requirements for our renewable energy project. Deep learning networks (DLNs) have been employed to obtain the correlations between meteorological features and power generation using a multilayer neural convolutional architecture with gradient descent algorithms to minimize estimation errors. This has wide applicability to the field of wind power forecasting. Therefore, this study aimed at the long-term (24–72-h ahead) prediction of wind power with an MAPE of less than 10% by using the Temporal Convolutional Network (TCN) algorithm of DLNs. In our experiment, we performed TCN model pretraining using historical weather data and the power generation outputs of a wind turbine from a Scada wind power plant in Turkey. The experimental results indicated an MAPE of 5.13% for 72-h wind power prediction, which is adequate within the constraints of our project. Finally, we compared the performance of four DLN-based prediction models for power forecasting, namely, the TCN, long short-term memory (LSTM), recurrent neural network (RNN), and gated recurrence unit (GRU) models. We validated that the TCN outperforms the other three models for wind power prediction in terms of data input volume, stability of error reduction, and forecast accuracy.
... Therefore, population size must be higher than three NP>3. The algorithm begins with a randomly initiated population which utilize NP D-dimension parameter vector within constrained by the minimum and maximum bounds (Ozer, 2010;Storn, & Price, 1997;Mandal, Chatterjee, & Bhattacharjee, 2013;Li, & Yin, 2012). ...
... Creation of the initial population composed of chromosomes with NP number and D size is calculated by using the equation ( used as a value, which varies between 0 and 2. This weighted difference chromosome is summed with the third chromosome (Ozer, 2010;Storn, & Price, 1997;Mandal, Chatterjee, & Bhattacharjee, 2013;Li, & Yin, 2012). 1,2,3, , , r NP r r r i . ...
... j=jrand condition is used for ensuring that at least one gene is obtained from the newly created chromosome. Gene, which is located in random j=jrand is selected from nj,i,g+1 irrespective of CR value (Ozer, 2010;Storn, & Price, 1997;Mandal, Chatterjee, & Bhattacharjee, 2013;Li, & Yin, 2012 ...
Machine Learning, Energy and Industrial Applications in Technology and Engineering Sciences / Chapter 6
https://iksadyayinevi.com/wp-content/uploads/2021/08/MACHINE-LEARNING-ENERGY-AND-INDUSTRIAL-APPLICATIONS-IN-TECHNOLOGY-AND-ENGINEERING-SCIENCES-.pdf
