The optimized portfolio formation is one of the important decisions making for corporations, accordingly a portfolio selection with top efficiency rate and controlled risk is one of problems that scholars attend them. In this research, we submit a method with multi objectives genetic algorithm based to portfolio formation and we lionize the value at risk as a paragon for risk measuring. Also we use 50 top companies’ data of stock exchange in time period from 1385 to 1389.The results show that multi objectives genetic algorithm can used to optimized portfolio formation and designed portfolio operation via genetic algorithm is different from with 50 top companies operation with equal weights.

Environmental problems and global warming are one of the biggest concerns of societies. Using renewable energy generation technologies as one of the most basic solutions is one of the main concerns of planners and beneficiaries of power grids, considering the variability of their output power day and night and their dependence on weather conditions. The uncertainty caused by these generations can have many effects on the costs imposed on the grid and the operation of electricity grids, such as an increase in power outages and energy not supplied. To solve this problem, a comprehensive multi-objective and probabilistic model has been proposed to determine the installation location, type, and optimal capacity of DGs in the modern supply chain of electricity. The final objective of this model is to minimize energy losses, investment and operation costs, energy not supplied, and environmental emissions. The proposed methods have been implemented by MATLAB software on the Garver power grid and IEEE 33-bus distribution grid and solved by the multi-objective NSGA-II. The final model can be effectively used to plan the supply chain of the modern electricity grid with the influence of renewable energy-based products in various economic, environmental, and social dimensions.

Background and Objectives: The groundwater level of Silakhor plain has been decreasing significantly with the occurrence of successive droughts, industrial growth and increasing water needs. In addition, the cropping pattern of the region in recent years has led to the cultivation of water crops, which sets the need for efficient management in the allocation of limited water resources in the region. In this study, to determine the optimal cropping pattern of major crops in Silakhor plain, with the aim of maximizing farmers' incomes and available water and land constraints, two approaches using Linear Programming and using Multi-Objective Meta Heuristic Algorithms in different exploitation scenarios have been investigated. Materials and Methods: Before using Linear Programming and optimization algorithms, in the first step, 100 different exploitation scenarios with equal intervals were determined for each crop year. Rainfall of the last 10 years in monthly and seasonal conditions was modeled using Artificial Neural Network and Genetic Programming and a better model according to the evaluation criteria of modeling. Then the rainfall of the next three crop years was forecasted and the resulting nutrition was estimated. Due to the need for proper exploitation from aquifers, it is necessary to have less exploitation than recharging in the coming years. In Silakhor plain, 50% of groundwater abstraction is used for horticultural, industrial and drinking products. Therefore, 45% of the feeding volume in each crop year was considered as the minimum exploitation and 140% of the exploitation in 2015 was considered as the maximum exploitation. One approach to solving constrained problems using metaheuristic algorithms is to constrain the problem using the penalty function and define the minimization of the penalty function as a goal. In this regard, in the second step, using Linear Programming, the optimal cropping pattern was followed, by the maximum income of farmers with limited water exploitationin each scenario and available land. Then, by defining the mentioned limitations as different penalty functions, the unresolved issue and maximizing the farmers' income function was considered as the first goal and minimizing the penalty functions as the second goal. The multi-objective optimization algorithm continues to operate until the response obtained from Linear Programming is reached with a maximum error of one percent,It is also an acceptable answer that the amount of the fine is zero. In other words, the answer in question must not exceed the defined limits. In this study, the performance of three types of static, dynamic and classified dynamics penalty functions in three multi-objective algorithms NSGA-II, SPEA-II and PESA-II. Are evaluated. The following equation shows the general form of the objective functions. Cost Function (I): Maximum Net Income Cost Function (II): Minimum Penalty Functions. Results: The results show that along with increasing groundwater exploitation, farmers' incomes also increase,However, in the exploitation of more than 223. 5, 222. 2 and 225. 1 million cubic meters for the cropping years 2020-2021, 2021-2022 and 2022-2023, respectively, the limitation of the total arable land in Silakhor plain prevents the increase of crop cultivation. As a result, the income of farmers in the region will not change. The results of the algorithms also show that the best performance among the algorithms in this issue belongs to the SPEA-II, PESA-II and NSGA-II algorithms with the number of iterations of 12. 1, 14. 5 and 17. 8, respectively. Among the penalty functions, on average in all three algorithms, the best performance belongs to the classified dynamics, dynamic and static penalty functions with the number of iterations of 13. 1, 13. 7 and 17. 5, respectively. Conclusion: Due to the decrease of groundwater level in Silakhor plain, determining different scenarios of groundwater abstraction and optimizing the cropping pattern appropriate to each scenario, in addition to increasing the economic productivity of the region, also facilitates the management of water resources. It is impossible to introduce a single algorithm and penalty function to solve all optimization problems. However, based on the results of this study, to solve the linear constraint problems, the use of the SPEA-II algorithm with a classified dynamics penalty function is recommended.

In this paper the problem of multidisciplinary and multi-objective conceptual design optimization of an air launched projectile (ALP) is investigated. The proposed task is performed using a three degree of freedom (DOF) flight dynamics simulation model and taking into account all the constraints involved in the optimization process. To maximize the payload weight as well as the range of ALP, the vehicle weight and balance, aerodynamics and stability disciplines are selected in the optimization process using a model with moderate levels of fidelity for each subject. The ALP design optimization problem contains 14 design variables and 2 target functions that include the payload weight and the vehicle range. Finally, a performance based comparison of results between the optimized ALP and its non-optimum initial configuration has been made. In addition, a Monte Carlo analysis is performed over the optimal ALP design to see the effects of launching uncertainties in meeting the mission requirements.

Face recognition from digital images is used for surveillance and authentication in cities, organizations, and personal devices. Internet of Things (IoT)-powered face recognition systems use multiple sensors and one or more servers to process data. All sensor data from initial methods was sent to the central server for processing, raising concerns about sensitive data disclosure. The main concern was that all data from all sectors that could contain confidential information was placed in a central server. Federated learning can solve this problem by using several local model training servers for each region and a central aggregation server to form a global model in IoT networks. This article presents a novel approach to optimize data transfer and convergence time in federated learning for a face recognition task using Non-dominated Sorting Genetic Algorithm II (NSGA II). The aim of the study is to balance the trade-off between training time and model accuracy in a federated learning environment. The results demonstrate the effectiveness of the proposed approach in reducing data transfer and convergence time, leading to improved performance in face recognition accuracy. This research provides insights for researchers and practitioners to enhance the efficiency of federated learning in real-world applications.

Bone drilling process is one of the most common processes in the orthopedic surgeries and bone break treatments. It is also very frequent in dentistry and bone sampling operations. Bone is a complex material and the machining process itself is sensitive, so bone drilling is one of the most important, common and sensitive processes in Biomedical Engineering field. Orthopedic surgeries can be improved using robotic bone drilling systems and mechatronic bone drilling tools. In the present study, multiobjective optimization is performed on the temperature and trust force at two steps. At the first step, two regression models are developed for modeling the temperature and force in bone drilling process considering three design variables, namely tool’ s rotational speed (V), feed rate (f) and tool diameter (D). At the second step, using the regression models, multi-objective genetic algorithm is used for the Pareto based optimization of bone drilling process considering two conflicting objectives: temperature and force. It has been found out that there are considerable connections and feasible principles for an optimal design of the process in case of applying Pareto-based multi-objective optimization; otherwise, these interesting results would not be discernible.

Cloud computing is a model for convenient on-demand user’ s access to changeable and configurable computing resources such as networks, servers, storage, applications, and services with minimal management of resources and service provider interaction. Task scheduling is regarded as a fundamental issue in cloud computing which aims at distributing the load on the different resources of a distributed system in order to optimize resource utilization and response time. In this paper, an optimization-based method for task scheduling is presented in order to improve the efficiency of cloud computing. In the proposed approach, three criteria for scheduling, including the task execution time, the task transfer time, and the cost of task execution have been considered. Our method not only reduces the execution time of the overall tasks but also minimizes the maximum time required for task execution. We employ the Multi-objective Non-dominated Sorting Genetic Algorithm (NSGA-II) for solving the scheduling problem. To evaluate the efficiency of the proposed method, a real cloud environment is simulated, and a similar method based on Multi-Objective Particle Swarm Optimization is applied. Experimental results show the superiority of our approach over the baseline technique.

Control of reactive distillation (RD) column is a challenging task due to its high degree of non-linearity, strong interactions, steady state multiplicity, time delay, process uncertainties, and the large number of possible control configurations. On the other hand, important products such as methyl tertiary butyl ether and ether tertiary butyl are produced via this method. An optimal control of methyl tertiary butyl ether (MTBE) column is studied in this paper utilizing the Multiobjective Genetic Algorithm concept in conjunction of Proportional-Integral-Derivative (PID) controller. The novelty of the work is in the optimal tuning of PID controllers by minimizing of two objective functions (Overshoot and Integral of Absolute Error (IAE)) through Non-Dominated Sorting Genetic Algorithm-II (NSGA-II). Numerical results show that NSGA-II based tuning method has excellent ability in optimal control of MTBE RD column.