A New Hybrid Ensemble Deep Learning Model for Train Axle Temperature Short Term Forecasting

The history matching, a key step in the reservoir numerical simulation, is performed to obtain a set of appropriate reservoir parameters to match the modeling results with the oilfield historical production data. The history matching problem is solved conventionally with Gauss-Newton, L-M, BFGS, and LBFGS methods, which require calculation of the gradient of the objective function, and is very slow in solving large-scale problems. In recent years, modern optimization algorithms, such as the genetic algorithm, particle swarm optimization, and differential evolution algorithm, have been applied in the reservoir history matching. In this paper, the hybrid particle swarm optimization and gravitational search algorithm (PSOGSA) is used to carry out the research on the automatic reservoir history matching. First, the basic concepts of the particle swarm optimization (PSO) and the gravity search algorithm (GSA) were introduced. Then, the concept and algorithm of PSOGSA were described. Moreover, the actual oil field data were used to establish a geological model and an objective function, and the PSOGSA was used to carry out the automatic history matching. The PSOGSA was compared with other algorithms. Finally, the effects of parameters on the PSOGSA were analyzed. The results show that the PSOGSA has a good performance in the automatic reservoir history matching, increases the diversity of particles, improves the global search capability, and accelerates the convergence. The PSOGSA is superior to the PSO or the GSA.

— In this paper, two optimization algorithms i.e., Particle Swarm Optimization (PSO) and Gravitational Search Algorithm (GSA) are combined to produce a hybrid population-based algorithm PSOGSA. The approach of HPSOGSA is to use the assessment of PSO using the social thinking capability and the manipulation of GSA using the local search proficiency. To increase the performance of HPSOGSA, some benchmark functions are used. Two standard database ORL and YALE is used to assess the classification performance of the proposed method. Classification accuracy is compared for diverse number of training samples per class. Further lead of proposed method is confirmed by analyzing percentage improvement in classification accuracy. With limited accessibility of training sample, percentage enhancement is very successful for face recognition applications. Keywords: Particle Swarm Optimization (PSO), Gravitational Search Algorithm (GSA), Particle Swarm Optimization and Gravitational Search Algorithm (PSOGSA), Hybrid Particle Swarm Optimization and Gravitational Search Algorithm (HPSOGSA). Cite this Article Purvaa Saxena, Gargi Mishra, Apoorva Aggarwal, Ayush Priyadarshi, Nitin Kumar, Rahul. Pattern Recognition Using Hybrid Meta-Heuristic. Journal of Artificial Intelligence Research & Advances. 2019; 6(2): 100–108p.

This paper presents the development and application of hybrid particle swarm optimization and gravitational search algorithm (PSOGSA) using cooperative technique (CPSOGSA). In this work, the algorithm is used for functions optimization problem. The CPSOGSA is developed with multiple groups using master-slave architecture. Six benchmark functions were tested to verify the effectiveness of the optimization algorithm. The developed algorithm was compared with the existing PSO, GSA and hybrid PSOGSA. The results indicate that the proposed technique produces better optimization solution compared to the conventional optimization algorithms.

In the renewable energy sector, the extraction of parameters for solar photovoltaic (PV) cells is a widely studied area of research. Parameter extraction is a non-linear complex optimization problem for solar PV cells. In this research work, the authors have implemented the Tunicate swarm algorithm (TSA) to estimate the optimized value of the unknown parameters of a PV cell/module under standard temperature conditions. The simulation results have been compared with four different, pre-existing optimization algorithms: gravitational search algorithm (GSA), a hybrid of particle swarm optimization and gravitational search algorithm (PSOGSA), sine cosine (SCA), and whale optimization (WOA). The comparison of results broadly demonstrates that the TSA algorithm outperforms the existing optimization algorithms in terms of root mean square error (RMSE) and convergence rate. Furthermore, the statistical results confirm that the TSA algorithm is a better algorithm in terms of average robustness and precision. The Friedman ranking test is also carried out to demonstrate the competency and reliability of the implemented approach.

Hybridization of Constriction Coefficient Based Particle Swarm Optimization and Gravitational Search Algorithm for Function Optimization

During recent decades, using wind power as one of the most substantial renewable energy sources has been vastly intensified in power systems. Increasing wind power penetration in the power grid can have both detrimental and beneficial effects. Due to the uncertain nature of wind power, the system operator is faced with many challenges to make proper decisions for the electricity market. In this paper, a probabilistic method is proposed which not only effectively manages wind power sources along with other producers, but also takes the uncertain behavior of wind power in defined intervals into account. Using wind farm information in each hour, a new approach for calculating the probability density function (PDF) in uncertainty bounds is presented. By utilizing the calculated PDF two objective functions are proposed for minimizing the total expected cost during 24 h without and with considering consumers payments. The proposed objective functions have been minimized through a two-level optimization method using three different heuristic algorithms (Teaching-Learning-Based Optimization (TLBO), Hybrid Particle Swarm Optimization and Gravitational Search Algorithm (PSO-GSA), and Ant Lion Optimization (ALO)) on an 8-bus sample transmission network. The results show the total expected cost has been decreased up to 20% in some scenarios. In addition, the proposed method can satisfy consumers by keeping their payments close to the lowest amount according to the different scenarios.

Particle swarm optimization (PSO) has been proven to show good performance for solving various optimization problems. However, it tends to suffer from premature stagnation and loses exploration ability in the later evolution period when solving complex problems. This paper presents a sequential hybrid particle swarm optimization and gravitational search algorithm with dependent random coefficients called HPSO-GSA, which first incorporates the gravitational search algorithm (GSA) with the PSO by means of a sequential operating mode and then adopts three learning strategies in the hybridization process to overcome the aforementioned problem. Specifically, the particles in the HPSO-GSA enter into the PSO stage and update their velocities by adopting the dependent random coefficients strategy to enhance the exploration ability. Then, the GSA is incorporated into the PSO by using fixed iteration interval cycle or adaptive evolution stagnation cycle strategies when the swarm drops into local optimum and fails to improve their fitness. To evaluate the effectiveness and feasibility of the proposed HPSO-GSA, the simulations were conducted on benchmark test functions. The results reveal that the HPSO-GSA exhibits superior performance in terms of accuracy, reliability, and efficiency compared to PSO, GSA, and other recently developed hybrid variants.

The detection of variables that contribute to the variation of a system is one of the most important considerations in the industrial manufacturing processes. This work presents the combination of Mahalanobis–Taguchi system and a hybrid binary metaheuristic based on particle swarm optimization and gravitational search algorithm (BPSOGSA) to perform an optimal feature selection in order to detect the relevant variables in a real process of foam injection in automotive industry. The proposed method is compared with other feature selection approach based in binary PSO algorithm. The experimental results revealed that BPSOGSA is faster and successfully converge selecting a smallest subset of features than BPSO. Moreover, the feature selection effect is validated through other widely used machine learning algorithms which improve their accuracy performance when they are trained with the subset of detected variables by the proposed system.

Optimal static state estimation using improved particle swarm optimization and gravitational search algorithm

This paper presents a modified sliding mode controller (MSMC) for tracking purpose of electro-hydraulic actuator system with mismatched disturbance. The main contribution of this study is in attempting to find the optimal tuning of sliding surface parameters in the MSMC using a hybrid algorithm of particle swarm optimization (PSO) and gravitational search algorithms (GSA), in order to produce the best system performance and reduce the chattering effects. In this regard, Sum square error (SSE) has been used as the objective function of the hybrid algorithm. The performance was evaluated based on the tracking error identified between reference input and the system output. In addition, the efficiency of the designed controller was verified within a simulation environment under various values of external disturbances. Upon drawing a comparison of PSOGSA with PSO and GSA alone, it was learnt that the proposed controller MSMC, which had been integrated with PSOGSA was capable of performing more efficiently in trajectory control and was able to reduce the chattering effects of MSMC significantly compared to MSMC-PSO and MSMC-GSA, respectively when the highest external disturbance, 10500N being injected into the system’s actuator.

Estimation of transformer parameters from nameplate data by imperialist competitive and gravitational search algorithms

Hypersonic reentry trajectory planning by using hybrid fractional-order particle swarm optimization and gravitational search algorithm

Power producers in a deregulated electricity market require a suitable strategy for bidding in order to maximize their benefit. Therefore, individual power supplier bids strategically for selecting coefficients of bid to oppose the strategy of rivals bidding. The oppositional learning based gravitational search algorithm (OGSA) is used in this paper to solve an optimal strategy of bidding problem. Gravitational search algorithm (GSA) is encouraged from mass interactions and law of gravity. In OGSA, opposite agents provide better optimal solutions compared to that of random agent solutions. The utilization of opposite agents in the evaluation process of GSA enables enhanced exploration of the search space. This prevents trapping of the search in local optimal solution and improves rate of convergence in GSA. The proposed method is tested on IEEE 30-bus system. The obtained results suggest that OGSA gives better solution in terms of profit as compared to other approaches such as Monte Carlo, Genetic Algorithm, Particle Swarm Optimization and Gravitational Search Algorithm.

In one-dimensional (1D) geomechanical modeling, precise predictions are essential for ensuring wellbore stability, defining safe mud weight windows, and determining hydraulic fracturing pressure ranges. However, accurately estimating geomechanical parameters, especially in data-limited environments, remains a significant challenge. This study addresses these challenges by integrating advanced machine learning algorithms with well log data from two wells in one of Iran’s hydrocarbon fields. The dataset includes shear wave velocity (Vs) logs, static Young’s modulus (Est) samples, uniaxial compressive strength (UCS) samples, internal friction angle ( φ ) measurements, and pore pressure (PP) samples. To tackle the challenge of predicting Vs with high precision, three sophisticated algorithms were employed: Bidirectional Long Short-Term Memory (BiLSTM), BiLSTM+Transformer, and a hybrid model (CBiT) combining 1D Convolutional Neural Network (CNN), BiLSTM, and Transformer mechanisms. Feature selection using Pearson correlation and Mutual Information and SHapley Additive exPlanations (SHAP) analysis identified the most influential parameters for each algorithm. The models were evaluated against traditional empirical methods using statistical metrics, such as mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R2). The results revealed that the machine learning models significantly outperformed the empirical methods, with the hybrid CBiT model showing superior accuracy, robustness, and generalization capability, establishing a new benchmark in predictive performance for shear wave velocity due to its ability to capture complex nonlinear relationships and long-term dependencies. Particle Swarm Optimization (PSO) was applied to further optimize model hyperparameters, achieving R2 values of 0.9989 (training) and 0.9985 (testing), with RMSE for unseen data reduced to 0.0269 and R2 reaching 0.9973. Dynamic elastic parameters, including Young’s modulus, bulk modulus, and shear modulus, were calculated using the optimized model. Strong correlations between dynamic and static properties, such as dynamic Young’s modulus and UCS, were extrapolated across the entire depth range of both wells. The study also introduced a novel approach to estimate internal friction angle using Plumb’s correlation and laboratory data, alongside improved pore pressure predictions based on Eaton’s empirical method. Finally, vertical and horizontal stresses were calculated using poroelastic equations for comprehensive stress modeling. This study demonstrates how the integration of hybrid machine learning algorithms, advanced optimization techniques, and rigorous data analysis can significantly improve the accuracy, reliability, and practical applicability of 1D geomechanical models, providing a robust and efficient framework for subsurface analysis and offering substantial advantages over conventional approaches.

<strong class="journal-contentHeaderColor">Abstract.</strong> In this paper, we have developed the Matlab code for a weighted hybrid of particle swarm optimization (PSO) and gravitational search algorithm (GSA) known as wPSOGSA, GSA, and PSO algorithms to interpret one-dimensional magnetotelluric (MT) data for some corrupted and non-corrupted synthetic data, as well as two examples of MT field data over different geological terrains: (i) geothermal rich area, Island of Milos, Greece, and (ii) Southern Scotland due to the occurrence of a significantly high electrical conductivity anomaly under crust and upper mantle extending from the Midland Valley across the Southern Uplands into northern England. Even though the fact that many models provide a good fit in a large predefined search space, specific models do not fit well. As a result, we used a Bayesian statistical technique to construct and assess the posterior probability density function (PDF) rather than picking the global model based on the lowest misfit error. This is proceeded by 68.27 % confidence interval for selecting a region where PDF is more prevalent to estimate the mean model which is more accurate and close to the true model. For illustration, correlation matrices show a significant relationship among layer parameters. The findings indicate, the wPSOGSA is less sensitive to model parameters and produces well, more stable and reliable results with the least uncertainty in the model that is compatible with existing borehole samples. Furthermore, the present methods resolve two additional geologically significant layers, one highly conductive (less than 1.0 Ωm) and another resistive (300.0 Ωm) over the Island of Milos, Greece, characterized by alluvium and volcanic deposits, respectively, as corroborated by borehole stratigraphy.