International Journal of Optimization in Civil Engineering

PROBABILISTIC GUIDE-SELECTION PARTICLE SWARM OPTIMIZATION

H. Rahami

Optimization is a key tool for solving complex engineering problems. This research introduces a novel particle swarm optimization algorithm in which all particles have a probability of being selected as guide particles, while the likelihood of each particle influencing others is determined proportionally to its performance. In other words, unlike the classical PSO algorithm where only the best particle is chosen as the fixed guide in each iteration, every particle can independently select its own guide based on the performance of other particles. This approach appears to prevent premature convergence of particles and enhance the exploration capability of the algorithm. Additionally, a parameter has been defined and investigated in this algorithm to adjust the ratio of exploration to exploitation power, which can be initialized according to the complexity type of the problem. The performance of the proposed algorithm was first evaluated using a set of benchmark mathematical functions, which confirmed the high accuracy of the algorithm in finding optimal solutions. Then, several truss design problems were examined as real structural case studies, and the obtained results indicate that the proposed algorithm exhibits suitable and acceptable performance compared with other well-known algorithms.

A HYBRID DIFFERENTIAL EVOLUTION AND SWARM ALGORITHM FOR ENGINEERING OPTIMIZATION

M. Shahrouzi

Although metaheuristic algorithms are popular tools for global optimization, none of them is reported as the best for all problems. Hybridization is an advanced solution to overcome the shortcomings of individual methods by using the power points of the others. Here, a popular swarm intelligent algorithm with high explorative capability is combined with an exploitative operator of differential evolution and some dynamic parameter variation, as well as a greedy operator to enhance the search refinement. The proposed method is evaluated on a variety of engineering and constrained engineering problems, including the optimal design of Belleville Spring, pressure vessel, car side impact problem, and Morrow point dam. According to the results, considerable improvement is observed with respect to the standard particle swarm optimizer as well as competitive performance with a number of metaheuristic algorithms.

SEISMIC LIFE-CYCLE COST OPTIMIZATION OF REINFORCED CONCRETE FRAMES USING METAHEURISTIC ALGORITHMS AND NEURAL NETWORKS

S. Gholizadeh

This paper employs a hybrid approach that integrates a metaheuristic algorithm with a properly trained neural network (NN) to perform seismic life‑cycle cost optimization of reinforced concrete (RC) frames within the framework of performance‑based design. In the proposed hybrid methodology, the center of mass optimization (CMO) metaheuristic algorithm is used to explore the design space. Additionally, a properly trained NN model is employed to estimate the nonlinear seismic response of the RC frames in order to evaluate the design constraints and compute the life‑cycle cost during the optimization process within a reasonable computational time. The efficiency of the proposed hybrid methodology is assessed through two performance‑based design optimization case studies involving 5‑ and 10‑story RC frames. The numerical results demonstrate that the proposed approach is an effective tool for optimizing the life‑cycle cost of RC frames by substantially reducing the computational burden of the optimization process.

PERFORMANCE-BASED OPTIMIZATION OF STEEL FRAMES WITH CHEVRON LATERAL RESTRAINT SYSTEM AND SEMI-RIGID CONNECTIONS USING ADVANCED METAHEURISTIC OPTIMIZATION ALGORITHMS

K. Farzad

This study investigates the optimal design of steel frames with chevron bracing systems and semi-rigid connections using a performance-based design framework and metaheuristic optimization algorithms. Optimization effectively balances design performance and cost in structural engineering. The three algorithms employed were selected based on their proven application to similar optimization problems, enabling identification of the most suitable approach for the present case. Chevron bracing offers architectural benefits and enhances lateral stiffness and strength. However, unbalanced vertical forces from tension and compression braces under seismic loading require nonlinear analysis for reliable capacity estimation. To address this, pushover analyses with multiple lateral load patterns are performed to capture responses consistent with performance-based design principles. Connection behavior plays a decisive role in the global performance of steel frames. Conventional assumptions of rigid or pinned connections oversimplify reality and produce inaccurate predictions. In this study, semi-rigid connections are modeled with greater fidelity by incorporating column panel zones (CPZs) and gusset plate stiffness at bracing joints. CPZs significantly influence energy dissipation and deformation, while gusset plates may contribute up to 40% of connection rotational stiffness. Neglecting these effects can underestimate interstory drift and misrepresent hinge mechanisms. Results show that accounting for initial connection stiffness improves both accuracy and cost efficiency. For 10- and 15-story frames, structural cost were reduced by about 7%, underscoring the value of realistic connection modeling in optimal design.

RECONSTRUCTING HISTORICAL EARTHQUAKE DATA FOR IRAN USING A DEEP NEURAL NETWORK OPTIMIZED BY ECBO ALGORITHM

N. Khavaninzadeh

This study develops a synthetic earthquake catalog for Iran (1900–1963) using a deep neural network (DNN) optimized by the Enhanced Colliding Bodies Optimization (ECBO) algorithm. The model, trained on post-1964 instrumental data from the Iranian Seismological Center, incorporates spatial, temporal, and tectonic features to estimate earthquake magnitudes. Statistical indices (MAE = 0.0064; RMSE = 0.3748) and bootstrap uncertainty analysis (±0.18 M) confirm the model’s reliability. The generated catalog provides a data-driven basis for improving seismic hazard assessment and historical seismicity reconstruction across the Iranian plateau.

A COMPREHENSIVE REVIEW OF DATA-DRIVEN AND METAHEURISTIC OPTIMIZATION IN CONCRETE MIXTURE AND STRUCTURAL DESIGN

P. Hosseini

Concrete mixture design optimization has evolved from traditional trial-and-error approaches to sophisticated computational methods. This paper presents a comprehensive review of optimization techniques applied to concrete mixture proportioning, covering statistical methods (Response Surface Methodology, Taguchi method), particle packing models, machine learning algorithms (Artificial Neural Networks, Random Forest, XGBoost, Support Vector Regression), and metaheuristic optimization techniques (Particle Swarm Optimization, Genetic Algorithms, EVPS, SA-EVPS). The review synthesizes findings from over 180 published studies, with detailed analysis of recent advances in artificial intelligence applications for multi-objective optimization of mechanical properties, cost, workability, durability, environmental sustainability, and structural performance. Key findings indicate that ensemble machine learning methods achieve superior prediction accuracy (R² > 0.95) for compressive strength, while metaheuristic algorithms effectively handle multi-objective trade-offs generating Pareto frontiers. The review also identifies critical research gaps including the need for standardized datasets, interpretable AI models, integration of life cycle assessment, and field validation of optimization results. Recent developments in self-adaptive algorithms (SA-EVPS) demonstrate improved convergence and solution quality for both material and structural optimization problems.

DAMAGE DETECTION IN SHELL STRUCTURES BY USING TIME SERIES ANALYSIS AND TOPOLOGY OPTIMIZATION

S. M. Tavakkoli

This paper presents a method for detecting the location and severity of damage in shell structures. The method relies on extracting time-domain damage-sensitive features from vibrational responses and applying topology optimization. To achieve this, singular values are extracted from the Hankel matrix using singular value (SVD) decomposition and selected as damage-sensitive features. The damage detection problem is formulated as a topology optimization problem in which damage is modeled using the solid isotropic material with penalization (SIMP) method. Sensitivity analysis is carried out using the finite difference method to compute the derivatives of the objective function with respect to the design variables, thereby enabling efficient gradient-based optimization. The objective function is defined to minimize the differences between the singular values of the reference structure and those of the model. Abaqus software is used to perform dynamic finite element analysis of the shell model and to derive acceleration responses at selected nodes, which serve as sensor locations. The results from several numerical examples demonstrate the high capability of the proposed method in accurately identifying both the location and severity of damage.

CHAOTIC VPS-SRM ALGORITHMS FOR OPTIMUM DESIGN OF THE LARGE-SCALE STRUCTURES

A. Zaerreza

The VPS-SRM algorithm is an enhanced metaheuristic approach developed for structural optimization. While it demonstrates robust performance in structural design, its efficiency remains subject to improvement, especially when dealing with large-scale structural optimization problems. To address this, the present study introduces improved versions of the VPS-SRM by incorporating chaotic maps. The performance of these chaotic-based variants was evaluated through the optimization of large-scale structural problems, including a 3-bay 15-story frame, 520-bar double-layer grid, and 800-bar double-layer grid. The results indicate that the chaotic versions significantly outperform the original algorithm, providing superior structural designs with higher precision and enhanced statistical results. Statistical analysis via the Kruskal-Wallis test further confirms that the chaotic variants offer a substantial improvement over the standard VPS-SRM.