Optimization Techniques Research Articles

A Gaussian process based surrogate approach for the optimization of cylindrical targets

Simulating direct-drive inertial confinement experiments presents significant computational challenges, both due to the complexity of the codes required for such simulations and the substantial computational expense associated with target design studies. Machine learning models, and in particular, surrogate models, offer a solution by replacing simulation results with a simplified approximation. In this study, we apply surrogate modeling and optimization techniques that are well established in the existing literature to one-dimensional simulation data of a new cylindrical target design containing deuterium–tritium fuel. These models predict yields without the need for expensive simulations. We find that Bayesian optimization with Gaussian process surrogates enhances sampling efficiency in low-dimensional design spaces but becomes less efficient as dimensionality increases. Nonetheless, optimization routines within two-dimensional and five-dimensional design spaces can identify designs that maximize yield, while also aligning with established physical intuition. Optimization routines, which ignore constraints on hydrodynamic instability growth, are shown to lead to unstable designs in 2D, resulting in yield loss. However, routines that utilize 1D simulations and impose constraints on the in-flight aspect ratio converge on novel cylindrical target designs that are stable against hydrodynamic instability growth in 2D and achieve high yield.

A comparative study of machine learning and response surface methodologies for optimizing wear parameters of ECAP-processed ZX30 alloy

Magnesium, valued for its lightweight, recyclability, and biocompatibility, faces challenges like its poor wear behavior and mechanical properties that limit its adaptation for a multitude of applications. In this study, various statistical analyses, and machine learning (ML) techniques were employed to optimize equal channel angular pressing (ECAP) process parameters for improving the wear behavior of Mg-3wt.% Zn-0.7 wt% Ca alloy. ECAP was conducted up to four passes via route Bc at 250 °C. Wear testing of both as-annealed (AA) and ECAP-processed alloys was performed using the dry ball-on-flat wear method under varying loads, speeds, and time. One pass (1P) and 4Bc-ECAP resulted in a notable uniform grain refinement of 86 % and 91 %, respectively, compared to the AA. X-ray diffraction (XRD) analysis confirmed a refined structure attributed to extensive dynamic recrystallization. Mechanical wear testing revealed a significant reduction in volume loss (VL), up to 56 % and 28.5 % after 1P and 4Bc samples, respectively, compared to the AA sample, supported by the observed texture intensity. The coefficient of friction (COF) stabilizes at 0.30–0.45, indicating low friction characteristics. Next, by adjusting wear load and speed through design of experiments (DOE), the wear output responses, VL and COF, were experimentally investigated. The output responses were predicted in the second step using ML, 3D response surface plots, and statistical analysis of variance (ANOVA). According to the regression model, the minimal VL was attained at a 5 N applied load. Also, the wear speed and VL at different passes are inversely proportional. On the other hand, the optimal COF was obtained at applied load about 2–3 N and 250 mm/s at different passes. The wear process variables were then optimized using different optimization techniques namely, genetic algorithm (GA), hybrid DOE–GA, and multi-objective genetic algorithm (MOGA) approaches.

Enhancing residential heating load prediction with advanced machine learning and optimization techniques

The effective prediction of heating load in residential buildings is crucial for optimizing energy consumption and enhancing thermal comfort. This study investigates three prominent regression techniques Extra Tree Regression, Decision Tree Regression, and Support Vector Regression utilized for this purpose. To improve their predictive performance, a novel metaheuristic optimization algorithm, the Tyrannosaurus Rex Optimization Algorithm, is employed. The dataset used includes features such as Relative Compactness, Surface Area, Wall Area, Roof Area, Overall Height, Orientation, Glazing Area, and Glazing Area Distribution. Each machine learning model (Extra Tree Regression, Decision Tree Regression, Support Vector Regression) is trained on historical heating load data and optimized using the Tyrannosaurus Rex Optimization Algorithm to adjust internal parameters. Comparative analysis demonstrates that the proposed Extra Tree + Tyrannosaurus Rex Optimization Algorithm model achieves superior performance with an R-squared value of 0.987 and a low Root Mean Square Error of 1.167. By leveraging the strengths of different machine learning algorithms, this research presents a reliable approach for accurately predicting heating load in residential buildings. This method holds promise for improving energy efficiency, enhancing thermal comfort, and promoting sustainable building design.

Stacking sequence optimization of composite laminates for maximum fundamental frequency using Bayesian optimization computational framework

A computational framework combined with the commercial finite element software Abaqus and Bayesian optimization algorithm is proposed. The proposed computational framework leverages the Gaussian process based-probabilistic capability in Bayesian optimization as a surrogate model for minimizing the computational cost of objective function evaluations in the finite element analysis. The optimization problem in this work is to enhance the maximum fundamental frequency of the composite laminates, which is one of the critical parameters in the design for composite structures. The optimization of stacking sequence selection is investigated and validated by the results obtained from literatures. The effectiveness and efficiency of the proposed Bayesian optimization computational framework in maximizing the fundamental frequency of composite laminates are demonstrated by comparing the optimized results from different optimization techniques through a series of cases including the 8-layer rectangular plates with 11 different boundary conditions and the 10 to 20-layer trapezoidal plates with the same boundary condition. The proposed framework developed in this work has highly potential as an engineering tool to address a broader range of structural optimization in vibration problems.

Synthesis of concentric circular antenna array for reducing the sidelobe level by employing sine cosine optimization algorithm

AbstractThe sine cosine algorithm (SCA), a meta‐heuristic optimization method, is used in this study to provide a precise linear and elliptical antenna array design for synthesizing the ideal far‐field radiation pattern in the fifth‐generation (5G) communication spectrum. The wireless communication system will undergo dramatic changes thanks to the forthcoming 5G technology, which offers exceptionally high data rates, increased capacity, reduced latency, and outstanding service quality. The most important component of 5G communications is an accurate antenna array design for an optimum far‐field radiation pattern synthesis with a suppressed sidelobe level (SLL) value and half power beam width (HPBW). While long‐distance communication necessitates a low HPBW, the entire side lobe area needs a suppressed SLL to prevent interference. The SCA is used in this case to the optimal feeding currents applied to each array member in the design examples of the concentric circular antenna arrays (CCAA) discussed in this article. It shows the litheness and attainment of the propound algorithm named SCA, chosen CCAAs with three rings and varying amounts of components or antenna array sets those are stated as follows: Set I (4, 6, 8 elements), Set II (8, 10, 12 elements), Set III (6, 12, 18 elements), Set IV (8, 14, 20 elements) with and without the center element are amalgamate. Apply the PSO, Jaya, and SCA optimization algorithms for all four Sets of antenna arrays and compare the attained results; the SLL values achieved by the SCA technique are contrasted with those of other current optimization techniques. The outcomes of all examinations reveal that the SCA algorithm achieved a superior SLL reduction over other optimization techniques.

UAV path planning techniques: a survey

Unmanned Aerial Vehicles (UAVs) are ideally suited for many real-world applications ranging from scientific to commercial, industrial, and military fields. Enhancing the efficiency of UAV-based missions through optimization techniques is of paramount significance. In this regard, the path planning problem that refers to finding the best collision-free path between the start point and the destination by addressing temporal, physical, and geometric constraints is a key issue. In this paper, a review of recent path planning methods from different perspectives with a clear and comprehensive categorization is presented. This study provides a general taxonomy categorizing the existing works into classical approaches, soft-computing techniques, and hybrid methods. Here, a detailed analysis of the recent techniques as well as their advantages and limitations is offered. Additionally, it provides an overview of environment modeling methods, path structures, optimality criteria, completeness criteria, and current UAV simulators.

Numerical solutions to Hyperbolic Maxwell quasi-variational inequalities in Bean–Kim model for type-II superconductivity

This paper is devoted to the finite element analysis for the Bean–Kim model governed by the full 3D Maxwell equations. Describing type-II superconductivity at the macroscopic level, this model leads to a challenging coupled system consisting of the Faraday equation and a hyperbolic quasi-variational inequality (QVI) of the second kind with L1-type nonlinearity, that arises explicitly from the magnetic field dependency in the critical current. With the involved Maxwell coupling in the 3D H(curl)-setting, the hyperbolic QVI character poses the primary challenge in the numerical investigation. Two mixed finite element methods based on implicit Euler and leapfrog time-stepping are proposed. On the one hand, the implicit Euler method results in a nonstandard system of curl-curl elliptic QVI with a first-order curl-type nonlinearity. Though the well-posedness of this system is guaranteed, its numerical realization is not straightforward and requires the use of a two-stage iteration process of high computational complexity. On the other hand, by approximating the electric and magnetic fields at two different time step levels, the leapfrog method turns out to be more suitable as it naturally eliminates the notorious QVI structure. More importantly, utilizing suited subdifferential and optimization techniques, we are able to prove an efficiently computable explicit formula for its exact solution in terms of the electric field, which makes its numerical computation substantially more favorable than the Euler method. As further advantages, the leapfrog method applies to broad scenarios involving low regular data of bounded variation (BV) in time for both the applied current source and the temperature distribution. Through nonstandard technical arguments tailored to the BV data, our analysis proves the conditional stability and, eventually, the uniform convergence of the proposed leapfrog method. This paper is closed by 3D numerical tests showcasing the reasonable and efficient performance of the proposed numerical solution.

Experimental investigation and optimization of performance, emission, and vibro-acoustic parameters of SI engine fueled with n-propanol and gasoline blends using ANN-GA coupled with NSGA3-modified TOPSIS hybrid approach

In the present study, performance, emission, and vibro-acoustic studies were conducted on a spark ignition (SI) engine fueled with gasoline and an n-propanol blend at variable compression ratio (CR), speed, and propanol blend fraction (PBF). Experimental data were used to model an artificial neural network (ANN) trained with a genetic algorithm (GA). ANN predictive responses were employed to establish regression relationships between brake power (BP), brake specific fuel consumption (BSFC), brake thermal efficiency (BTE), oxides of nitrogen (NOx), carbon monoxide (CO), hydrocarbon (HC), resultant vibration acceleration (RVA), and sound pressure level (SPL) with operating parameters using response surface methodology (RSM). These models served as objective functions in the non-dominated sorting genetic algorithm-3 (NSGA3), a multi-objective optimization (MOO) technique, to optimize responses and obtain non-dominated solutions. These solutions were filtered using a modified technique for order preference by similarity to the ideal solution (TOPSIS) to obtain a compromised optimal solution. ANN-GA model outcomes showed high accuracy, with coefficient of determination (R2) and root mean square error (RMSE) values ranging from 0.979 to 0.993 and 0.0381 to 0.0643, respectively. NSGA3 coupled with modified TOPSIS identified optimal operating conditions at 1271.77 RPM, a CR of 11.96, and a PBF of 33.26 %.

Use of Muscle Relaxants for Acute Postoperative Pain: A Practical Review.

Skeletal muscle relaxants have emerged as a broad category of analgesic adjuncts, aiming to improve pain relief and reduce opioid reliance. These drugs induce muscle relaxation and reduce muscle spasms, and therefore, they are commonly used in surgical procedures involving muscle manipulation, such as abdominal surgery, breast surgery, and spine surgery. However, their analgesic efficacy and opioid-sparing effects are poorly explored. A scoping review of literature was performed with several electronic databases. We used a search string with a sequence of text words and word variants related to central muscle relaxants, pain management, postoperative pain, and specific muscle relaxants. Review of literature shows significant heterogeneity among the studies in terms of surgical procedures, patient populations, choice of muscle relaxant, and timing and duration of administration, potentially limiting the generalizability of the findings and the ability for pooled analysis. Given the lack of evidence, we recommend that the use of skeletal muscle relaxants be reserved for patients in whom an optimal multimodal analgesic technique is not adequate. Also, there may be a limited role for these drugs in patients at high risk of postoperative pain undergoing surgical procedures with expected high opioid requirements. Due to the concerns of potential adverse effects, the decision to use muscle relaxants in vulnerable populations should be made carefully, weighing the benefits against the risks.

Optimized backstepping‐based finite‐time containment control for nonlinear multi‐agent systems with prescribed performance

AbstractIn this article, a finite‐time optimal containment control method is proposed for nonlinear multi‐agent systems with prescribed performance. First, a neural network‐based reinforcement learning algorithm is developed under the optimized backstepping framework. The algorithm employs an identifier‐critic‐actor architecture, where the identifiers, critics and actors are used to estimate the unknown dynamics, evaluate the system performance, and optimize the system, respectively. Subsequently, in order to guarantee the transient performance of the tracking error, the original system is converted into an equivalent unconstrained system. Then, the tracking errors are allowed to converge to a prescribed set of residuals in finite time by combining prescribed performance control and finite‐time optimal control techniques. Furthermore, by using the Lyapunov stability theorem, it is verified that all signals are semi‐globally practical finite‐time stable, and all followers can converge to a convex region formed by multiple leaders. Finally, the effectiveness of the proposed scheme is demonstrated by a practical example.

Optimal phase-selective entrainment of electrochemical oscillators with different phase response curves.

We investigate the entrainment of electrochemical oscillators with different phase response curves (PRCs) using a global signal: the goal is to achieve the desired phase configuration using a minimum-power waveform. Establishing the desired phase relationships in a highly nonlinear networked system exhibiting significant heterogeneities, such as different conditions or parameters for the oscillators, presents a considerable challenge because different units respond differently to the common global entraining signal. In this work, we apply an optimal phase-selective entrainment technique in both a kinetic model and experiments involving electrochemical oscillators in achieving phase synchronized states. We estimate the PRCs of the oscillators at different circuit potentials and external resistance, and entrain pairs and small sets of four oscillators in various phase configurations. We show that for small PRC variations, phase assignment can be achieved using an averaged PRC in the control design. However, when the PRCs are sufficiently different, individual PRCs are needed to entrain the system with the expected phase relationships. The results show that oscillator assemblies with heterogeneous PRCs can be effectively entrained to desired phase configurations in practical settings. These findings open new avenues to applications in biological and engineered oscillator systems where synchronization patterns are essential for system performance.

LUCMT: Learnable Under-sampling and Reconstructed Network with Cross Multi-head Attention Transformer for Accelerating MR Image Reconstruction

Background and ObjectiveAs a widely used technique for Magnetic Resonance Image (MRI) acceleration, compressed sensing MRI involves two main issues: designing an effective sampling strategy and reconstructing the image from significantly under-sampled K-space data. In this paper, an innovative approach is proposed to address these two challenges simultaneously. MethodsA novel MRI reconstruction method, termed as LUCMT, is implemented by integrating a learnable under-sampling strategy with a reconstruction network based on the Cross Multi-head Attention Transformer. In contrast to conventional static sampling methods, the proposed adaptive sampling scheme is processed optimally by learning the optimal sampling technique, which involves binarizing the sampling pattern by a sigmoid function and computing gradients by backpropagation. And the reconstruction network is designed by using CS-MRI depth unfolding network that incorporates a Cross Multi-head Attention (CMA) module with inertial and gradient descent terms. ResultsT1 brain MR images from the FastMRI dataset are used to validate the performance of the proposed method. A series of experiments are conducted to validate the superior performance of our proposed network in terms of quantitative metrics and visual quality. Compared with other state-of-the-art reconstruction methods, LUCMT achieves better reconstruction performances with more accurate details. Specifically, LUCMT achieves PSNR and SSIM results of 41.87/0.9749, 46.64/0.9868, 50.41/0.9924, and 53.51/0.9955 at sampling rates of 10%, 20%, 30%, and 40%, respectively. ConclusionsThe proposed LUCMT method can provide a promising way for generating optimal under-sampling mask and accelerating MRI reconstruction accurately.

A comparative study of remotely sensed reservoir monitoring across multiple land cover types

Continuous monitoring of reservoirs and dams is essential for efficient water management. Synthetic Aperture Radar (SAR) imagery offers the potential for continuous monitoring of surface water through all-weather ground observation. The objective of this study is to enhance the accuracy of water body detection and water quantity estimation by applying 64 combinations of speckle filtering and object detection techniques to Sentinel-1 imagery. For speckle filtering, the Median, Gaussian, Lee, and Frost techniques were used with various window sizes (3, 5, 7, and 9). For water body detection, the Otsu, Kittler-Illingworth (KI), Chan-Vese (CV), and K-means methods were employed. The study area included three reservoirs and two dams in Korea, encompassing a variety of water surface sizes and types of land cover. To validate the accuracy of each water body detection combination, manual delineation-based water mask images from Sentinel-2 were employed. Furthermore, a regression equation (y=axb) between water surface area and storage was used to estimate water storage based on SAR imagery, followed by time-series validation using in-situ data. The research results indicate that the optimal detection technique varies significantly depending on the type of surrounding land cover and the size of the water body. The highest performance was observed for the CV technique combination for waterfront pixels, and for the KI technique combination for other land cover pixels. In speckle filtering techniques using a large window size, the false detection rate caused by vegetation and buildings was low; however, the boundaries of water bodies were blurred. Consequently, using smaller window sizes in SAR imagery and leveraging optimal water body detection combinations specific to land cover types, along with post-processing using masking data, would enhance the performance of water surface area and storage estimation.

Non-optimal and optimal fractional control analysis of measles using real data

This study employs fractional, non-optimal, and optimal control techniques to analyze measles transmission dynamics using real-world data. Thus, we develop a fractional-order compartmental model capturing measles transmission dynamics. We then formulate an optimal control problem to minimize the disease burden while considering constraints such as vaccination resources and intervention costs. The proposed model’s positivity, boundedness, measles reproduction number, and stability are obtained. The sensitivity analysis using the partial rank correlation coefficient is shown for the fractional orders of 0.99 and 0.90. It is noticed that the rate of recruitment into the susceptible population (π), the rate at which individuals in the latent class become asymptomatic (α1), and the transmission rate (β) contribute positively to the spread of the disease, while the rate at which individuals in the asymptomatic class become symptomatic (α2), the vaccination rate for the first measles dose (γ1), and the rate at which individuals in the latent class recover from measles (δ1) contribute significantly to the reduction of measles in the community. Utilizing numerical simulations and sensitivity analyses, we identify optimal control strategies that balance the trade-offs between intervention efficacy, resource allocation, and societal costs. Our findings provide insights into the effectiveness of fractional optimal control strategies in mitigating measles outbreaks and contribute to developing more robust and adaptive disease control policies in real-world scenarios.

Analysis and development of smart production and distribution line system in smart grid based on optimization techniques involving digital twin

The term Digital Twin (DT) is defined as the virtual demonstration of an object that is represented through real-time datasets. DT is done through artificial intelligence to enhance decision-making techniques. DT includes the process of simulation, amalgamation, observation, analysis, and conservation. The DT is simply the exact reproduction of the physical structures. DT is used in the identification and evaluation of problems through real-time analysis. It is important to have prior analysis and evaluation of the object before existing in the real world. These digital twins help in the manufacturing and implementation of the production line system. DT includes the production line with the station division and the hours needed for the operating conditions for the assembly process. The systems are integrated to reduce the overall cost parameter. The physical simulation model is employed to obtain higher performance with reduced cost. An artificial neural network with a genetic algorithm is used for the optimization process to achieve a production line system using digital twins.

Improving photovoltaic cell parameter calculations through a puffer fish inspired optimization technique

The precise estimation of solar PV cell parameters has become increasingly important as solar energy deployment expands. Due to the intricate and nonlinear characteristics of solar PV cells, meta-heuristic algorithms show greater promise than traditional ones for parameter estimation. This study utilizes the Puffer Fish (PF) meta-heuristic optimization method, inspired by male puffer fish's circular structures, to estimate parameters of a modified four-diode PV cell. The PF algorithm's performance is assessed against ten benchmark test functions, with results presented as mean and standard deviation for validation. Comparative analysis with Particle Swarm Optimization (PSO), Grey Wolf Optimization (GWO), Rat Search Algorithm (RAT), Heap Based Optimizer (HBO), and Cuckoo Search (CS) algorithms highlights PF's superior performance, achieving optimal solutions with minimal error of 7.8947E-08. Statistical tests, including Friedman Ranking (1st) and Wilcoxon's rank sum (3.8108E-07), confirm PF's superiority. The circular structures of male puffer fish serve as an effective model for optimization algorithms, enhancing parameter estimation. Benchmark tests and statistical analysis consistently underscore PF's superiority over other meta-heuristic algorithms. Future research should explore PF's potential applications in solar energy and beyond.

Group decision making method for third-party logistics management: An interval rough cloud optimization model

Group decision making in third-party logistics service selection plays an essential role for improving service quality, increasing efficiency and reducing the net cost. Fuzzy and uncertain linguistic variables are commonly used to represent experts‘rankings in optimization problems. To recognize the limits of human cognition and subjectivity of human evaluations, several optimization approaches have been studied to select remanufacturing alternatives in decision making processes, however these methods have certain deficiencies such as lacking manipulation tools of diverse information, randomness, use of predefined parameters increasing uncertainty, interpersonal relations among evaluation criteria. The integration of interval numbers, rough approximations, and cloud model theory plays a significant role to model incomplete and inadequate information occurring in decision making problems. This research paper focuses on the integration of dual interval rough integrated cloud model with best-worst optimization technique, Multi-Attributive Border Approximation area Comparison (MABAC) and Weighted Aggregated Sum Product Assessment (WASPAS) approaches. A novel min–max optimization model based dual interval rough integrated cloud values is designed to compute the weight coefficients and consistency ratio for each criteria. The consistency of proposed optimization model is checked using a consistency ratio test. Secondly, the alternatives are ranked using the proposed DIRI cloud based MABAC and WASPAS approaches using interval clouds based weighted sum and weighted product coefficients, approximation area values and distance formulae. The significance of the proposed model is highlighted with a case study of third-party logistics service management of an electronic firm. The rationality and out-performance of the proposed methodology is studied by a comparative analysis with existing approaches and detailed sensitivity analysis on different variations of criteria weights and parameter.

Optimal production plan in a production inventory model under the coexistence of both green and regular products: A study on Indian textile industry

The necessity to restrict environmental pollution worldwide has led to the development and implementation of green products. Various sectors within the industrial domain are now placing emphasis on the creation of environmentally friendly products in order to minimize their impact on the environment. The textile industry, both in India and globally, significantly contributes to environmental pollution through the release of untreated effluents into water and the emission of carbon into the atmosphere. Consequently, the popularity of green apparel is steadily increasing as a means to alleviate this crisis. It has been reported that, despite customer preference for green apparel, factors such as the cost of these products and a lack of consumer awareness regarding their importance deter a significant portion of potential buyers from making purchases. In order to stimulate demand for green apparel, the government could implement a subsidy scheme and initiate advertising campaigns highlighting the benefits of using such products. In the market, green apparel and regular apparel are viewed as substitutes for one another and coexist. As a result, there is a constant price competition between these two types of products. This research examines the optimal inventory decisions of manufacturers in a production inventory model where both green and regular products are present and product substitution is allowed. The demand for both types of products is influenced by their respective selling prices. Furthermore, the level of greenness and the efforts put into advertising green products have positive effects on generating demand for these products. The resulting objective function in this model is non-concave, and the quantum behaved particle swarm optimization technique is utilized to solve the corresponding optimization problem. Additionally, Mathematica software has also been employed to verify the efficiency of the proposed solution algorithm. To provide economic validation of the proposed model, a numerical example is presented. Finally, a sensitivity analysis of the key parameters is conducted to gain insights into their impact on the optimal solution. The results indicate that strategic and well-executed advertising campaigns, along with government subsidy schemes, can boost the market for green products and generate greater profits for manufacturers.

A novel application of artificial intelligence technology for outdoor high-voltage composite insulator

The distribution of the electric field plays a crucial role in evaluating the effectiveness of composite insulators. This research presents an innovative approach for optimizing the design of corona rings, aiming to decrease the electric field intensity in critical areas to satisfactory levels. The investigation examined the correlation among corona ring design variables and the intensity of the electric field close to the energized-end fitting in a 220 kV composite insulator equipped with a corona ring. Using design of experiment methods, the impact of three corona ring design parameters – corona ring diameter (R), ring tube diameter (Dr), and corona ring height (H) – was assessed. A novel nonlinear mathematical objective function was formulated, connecting the electric field magnitude to the structural parameters of the corona ring. Subsequently, the Bat algorithm, a bio-inspired optimization technique, was employed to enhance the initial corona ring design and mitigate the electric field intensity. Finally, the Finite Element Method (FEM) was utilized to simulate and evaluate the voltage distribution and electric field stress. The optimization process resulted in a substantial decrease in the highest electric field magnitude, by 58.6 % compared to the insulator with the threshold value, and 75 % when devoid of a corona ring. Additionally, the research highlighted the significant influence of ring tube thickness on the distribution of the electric field. The combination of experimental design and the Bat algorithm proved to be a powerful tool for optimizing the design of corona rings on transmission line composite insulators, providing precise solutions for optimizing corona discharge problems and enhancing the reliability of power transmission systems.

Method for layout optimization of coastal cage aquaculture systems in Southeast Asia

This paper presents the development of a generally applicable method for the layout optimization of coastal cage aquaculture systems in Southeast Asia. The method combines data from the authors’ spatial-planning methodology with results of water quality modelling and an optimization technique based on genetic algorithms. The effectiveness of the method is successfully demonstrated for a target aquaculture site in northwest Bali, Indonesia. Optimum siting and sizes of fish farms to maximize fish production and promote fish growth while securing environmental sustainability are delivered for the site in Bali. The results of the application to the western channel of the site show that by proper fish farming layout, environmentally sustainable fish production in the site in Bali could rise tremendously with respect to the current operating conditions. It is found that installing many smaller fish farms than fewer larger farms on the western tidal channel of the aquaculture site is much more profitable. The method has broad applicability and can be upscaled to new sites, and for assessment, expansion, and optimization of the currently operating sites.