Multi-objective Genetic Optimization Research Articles

Dynamic cellular manufacturing system considering machine failure and workload balance

Machines are a key element in the production system and their failure causes irreparable effects in terms of cost and time. In this paper, a new multi-objective mathematical model for dynamic cellular manufacturing system (DCMS) is provided with consideration of machine reliability and alternative process routes. In this dynamic model, we attempt to resolve the problem of integrated family (part/machine cell) formation as well as the operators’ assignment to the cells. The first objective minimizes the costs associated with the DCMS. The second objective optimizes the labor utilization and, finally, a minimum value of the variance of workload between different cells is obtained by the third objective function. Due to the NP-hard nature of the cellular manufacturing problem, the problem is initially validated by the GAMS software in small-sized problems, and then the model is solved by two well-known meta-heuristic methods including non-dominated sorting genetic algorithm and multi-objective particle swarm optimization in large-scaled problems. Finally, the results of the two algorithms are compared with respect to five different comparison metrics.

Shape memory alloy-based mechanism for aeronautical application: Theory, optimization and experiment

Efforts to create efficient and lighter aeronautical structures are defining morphing systems especially those associated with smart materials. In this regard, three simple mechanisms using shape memory alloy (SMA) wires are investigated to generate torque that could be used for flap actuation. The devices consist of an SMA wire biased by a linear spring in the following configurations: concurrent, collinear, and in parallel attached to a pulley. The design of such mechanisms are modeled, optimized, and experimentally verified. The model for the flap consists of two rigid bodies, one fixed and the other rotating, with a single actuator connected to each body. Aerodynamic loading and heat transfer analysis are also considered. The model utilizes the thermomechanical properties for an SMA wire experimentally characterized via improved inverse problem techniques. A multiobjective genetic optimization is implemented to find designs for the three configurations that minimize power consumption and maximize flap deflection magnitude. Overall, as design complexity (i.e., number of degrees of freedom) increases, the power to achieve a certain flap deflection decreases. The maximum deflection for all three mechanisms is sufficient for typical aircraft operations. Finally, numerical results were verified via an experimental apparatus, where similar performance to the model was achieved.

Scaling effects of a novel solar receiver for a micro gas-turbine based solar dish system

Laboratory-scale component testing in dedicated high-flux solar simulators is a crucial step in the development and scale-up of concentrating solar power plants. Due to different radiative boundary conditions available in high-flux solar simulators and full-scale power plants the temperature and stress profiles inside the investigated receivers differ between these two testing platforms. The main objective of this work is to present a systematic scaling methodology for solar receivers to guarantee that experiments performed in the controlled environment of high-flux solar simulators yield representative results when compared to full-scale tests. In this work the effects of scaling a solar air receiver from the integration into the OMSoP full-scale micro gas-turbine based solar dish system to the KTH high-flux solar simulator are investigated. Therefore, Monte Carlo ray-tracing routines of the solar dish concentrator and the solar simulator are developed and validated against experimental characterization results. The thermo-mechanical analysis of the solar receiver is based around a coupled CFD/FEM-analysis linked with stochastic heat source calculations in combination with ray-tracing routines. A genetic multi-objective optimization is performed to identify suitable receiver configurations for testing in the solar simulator which yield representative results compared to full-scale tests. The scaling quality is evaluated using a set of performance and scaling indicators. Based on the results a suitable receiver configuration is selected for further investigation and experimental evaluation in the KTH high-flux solar simulator.

A multi-variable multi-objective methodology for experimental data and thermodynamic analysis validation: An application to micro gas turbines

A methodology aimed at defining thermodynamic model parameters and validating experimental data has been proposed. The methodology consists of a thermodynamic model of a micro gas turbine coupled with a multi-variable multi-objective genetic optimization algorithm, in which decision variables and objectives are set depending on available experimental data. To validate both the thermodynamic model and the collected experimental data, the methodology has been applied to two micro gas turbine plants: the Capstone C30 and the Turbec T100. Validations of the thermodynamic model and the collected experimental data for the two plants have been performed by evaluating the match between input and output physical parameters. The optimal results of the optimization algorithm have plausible thermodynamic parameters and constitute the Pareto front; between these results, the one with the minimum difference between experimental data and calculated values is chosen as preferred. The two studied cases highlight the effect of measurement chain errors on experimental data reliability: the greater is the overall variance of the objectives, the lower is the accuracy of the experimental data. The effectiveness of proposed methodology has been verified for the Capstone C30 through the congruence of the design operating conditions on both the compressor and turbine maps.

Exploration of System Vulnerability in Naval Ship Concept Design

This paper presents a methodology and analysis tool to explore naval ship system vulnerability using a simplified preliminary arrangement and vulnerability (Pa&V) modeling approach in the concept and requirements exploration (C&RE) process at Virginia Tech. This C&RE process uses a multiobjective genetic optimization (MOGO) to explore the design space for nondominated ship design solutions based on design effectiveness, cost, and risk. The previous C&rE process evaluated ship characteristics and only intact system options to determine a design overall measure of effectiveness (OMOE) through the calculation of Measures of Performance (MOPs). Vulnerability is very dependent on system design, architecture, and spatial arrangement, and for simplicity these were not developed or calculated. The new C&RE uses a simplified Pa&V model so that an Overall Measure of Vulnerability (OMOV) can be calculated for each ship design using the new process and tools described in this paper. The OMOV is calculated by combining Vulnerability measure of performance scores across multiple ship mission options. It is then combined with the intact MOPs to calculate a more complete warfighting OMOE which can be used in the C&RE. 1. Introduction Typically, survivability analysis for naval ships is deferred until preliminary design or later. Many important design decisions regarding systems and system architecture are made in concept exploration and are difficult and costly to reverse. survivability is an important factor in assessing mission effectiveness, which is an important objective attribute in concept exploration and must be considered in these early decisions. However, several questions are raised when attempting to assess and consider survivability early:Is sufficient model detail and fidelity available during concept exploration to make a reasonable survivability assessment?If sufficient detail is available, does early survivability consideration impact the resulting design decisions?Are simplified, but traditional vulnerability models able to discriminate correctly between alternative designs, answer the right questions, provide important insight, and enable good decisions for moving forward into concept development and preliminary design?Are tools and resources available to consider survivability early or do we need a completely new approach and new tools?

Multi-Objective Optimal Design of Bearingless Switched Reluctance Motor Based on Multi-Objective Genetic Particle Swarm Optimizer

In recent decades, bearingless switched reluctance motors (BSRMs) have been proposed. However, few researchers focused on the optimal design of the BSRMs. In this paper, the multi-objective optimal design of BSRMs is investigated. At first, an analytical design model is derived from the mathematical model of the BSRMs. An initial design is calculated by the analytical design model. The electromagnetic performance is compared with calculation results from the finite-element method (FEM). Then, the objective functions, constraints, and decision variables are also determined. Corresponding sensitivity analysis of the decision variables is implemented. Besides, aiming at solving the optimization problem with disconnected, non-uniformly distributed Pareto front and multiple local optimums, a novel multi-objective genetic particle swarm optimizer (MOGPSO) is presented. The algorithm performance of the proposed MOGPSO is validated by solving the standard test functions. Then the proposed MOGPSO is applied for the optimal design of BSRMs. Optimization results solved by MOGPSO, conventional multi-objective particle swarm optimizer, and non-dominated sorting genetic algorithm II are compared and analyzed. Comparison results reveal that the proposed MOGPSO can achieve more non-dominated solutions in Pareto front and is particularly suitable for optimization of BSRMs. The final optimal design is selected from the obtained Pareto front. The electromagnetic performance is compared with the initial design and verified by the FEM. Verification results show that the optimal design of BSRMs based on the analytical design model and the proposed MOGPSO is feasible and effective.

New mathematical modeling for a location\u2013routing\u2013inventory problem in a multi-period closed-loop supply chain in a car industry

This paper studies a location–routing–inventory problem in a multi-period closed-loop supply chain with multiple suppliers, producers, distribution centers, customers, collection centers, recovery, and recycling centers. In this supply chain, centers are multiple levels, a price increase factor is considered for operational costs at centers, inventory and shortage (including lost sales and backlog) are allowed at production centers, arrival time of vehicles of each plant to its dedicated distribution centers and also departure from them are considered, in such a way that the sum of system costs and the sum of maximum time at each level should be minimized. The aforementioned problem is formulated in the form of a bi-objective nonlinear integer programming model. Due to the NP-hard nature of the problem, two meta-heuristics, namely, non-dominated sorting genetic algorithm (NSGA-II) and multi-objective particle swarm optimization (MOPSO), are used in large sizes. In addition, a Taguchi method is used to set the parameters of these algorithms to enhance their performance. To evaluate the efficiency of the proposed algorithms, the results for small-sized problems are compared with the results of the ε-constraint method. Finally, four measuring metrics, namely, the number of Pareto solutions, mean ideal distance, spacing metric, and quality metric, are used to compare NSGA-II and MOPSO.

Multiobjective optimization–based fault‐tolerant flight control system design

SummaryThe loss of measurements used for controller scheduling or envelope protection in modern flight control systems due to sensor failures leads to a challenging fault‐tolerant control law design problem. In this article, an approach to design such a robust fault‐tolerant control system, including full envelope protections using multiobjective optimization techniques, is proposed. The generic controller design and controller verification problems are derived and solved using novel multiobjective hybrid genetic optimization algorithms. These algorithms combine the multiobjective genetic search strategy with local, single‐objective optimization to improve convergence speed. The proposed strategies are applied to the design of a fault‐tolerant flight control system for a modern civil aircraft. The results of an industrial controller verification and validation campaign using an industrial benchmark simulator are reported.

Experimental investigation and multiresponse genetic optimization of drilling parameters for self-healing GFRP

Self-healing materials heal partially or completely when damage occurs to restore the functionality of the material. Self-healing technology has found its applications in sensors, supercapacitors, solar cells, fabrics etc. Self-healing GFRP specimens were fabricated by embedding microcapsules filled with dicyclopentadiene monomer in glass fiber reinforced epoxy resin. Drilling experiments were conducted on self-healing GFRP panels comprised of varied quantities and sizes of microcapsules under different drilling conditions to find the effect of adding microcapsules in the material on the drilling characteristics. Thrust force and temperature variations during drilling were studied, and it was found that that the microcapsule concentration in the composite had marked effect on both the measured outputs. Genetic multiobjective optimization was employed to obtain the optimum pareto values to determine the optimum thrust and temperature.

Multiobjective Optimization of Multistage Synchronous Induction Coilgun Based on NSGA-II

The structure and trigger control strategy have become the most important factors that restrict the performance of the multistage synchronous induction coilgun (MSSICG). However, it is still a difficult task to design MSSICG under overload constraint due to coupling between the multiple parameters. In this paper, the maximization of the emission efficiency and acceleration stationarity is treated as a multiobjective optimization problem. By analyzing the relationship between the number of turns and the other structural parameters of the launch, the multiobjective optimization model of MSSICG is established by the current filament method which was verified by the experimental data and finite-element method. And then the second generation nondominated sorting genetic algorithm (NSGA-II) and multiobjective particle swarm optimization (MOPSO) were employed to optimize the model in order to maximize the energy transfer efficiency while achieving the smooth acceleration of the armature. With the formulated optimization model, a five-stage synchronous induction coilgun is optimized as a special case. A decision-making procedure based on the fuzzy membership function is used for obtaining best compromise solution from the set of Pareto-solutions obtained through NSGA-II and MOPSO. In addition, the optimization performance of the proposed multiobjective optimization model and the single-objective optimization model of the MSSICG was compared. The result of optimization shows that the proposed multiobjective optimization model of MSSICG can effectively improve the performance of the coilgun compared with the single-objective optimization model which takes of the launch velocity and overload acceleration as the combination objective function or only the launch velocity.

Integrated modeling and solving the resource allocation problem and task scheduling in the cloud computing environment

Cloud computing is considered to be a new service provider technology for users and businesses. However, the cloud environment is facing a number of challenges. Resource allocation in a way that is optimum for users and cloud providers is difficult because of lack of data sharing between them. On the other hand, job scheduling is a basic issue and at the same time a big challenge in reaching high efficiency in the cloud computing environment. In this paper, “the cloud resources management problem” is investigated that includes allocation and scheduling of computing resources, such that providers achieve the high efficiency of resources and users receive their needed applications in an efficient manner and with minimum cost. For this purpose, a group technology based non-linear mathematical model is presented with an aim at minimization of load difference of servers, number of transfers between servers, number of active virtual machines, maximum construction time, the cost of performing jobs and active servers energy consumption. To solve the model, a meta-heuristic multi-objective hybrid Genetic and Particle Swarm Optimization algorithm is proposed for resource allocation and scheduling. In order to demonstrate the validity and efficiency of the algorithm, a number of problems with different dimensions are randomly created and accordingly the efficiency and convergence capability of the suggested algorithm is investigated. The results indicated that the proposed hybrid method has had an acceptable performance in generating high quality, diverse and sparse solutions.

Dynamic Self-Adaptive Double Population Particle Swarm Optimization Algorithm Based on Lorenz Equation

In order to improve some shortcomings of the standard particle swarm optimization algorithm, such as premature convergence and slow local search speed, a double population particle swarm optimization algorithm based on Lorenz equation and dynamic self-adaptive strategy is proposed. Chaotic sequences produced by Lorenz equation are used to tune the acceleration coefficients for the balance between exploration and exploitation, the dynamic self-adaptive inertia weight factor is used to accelerate the converging speed, and the double population purposes to enhance convergence accuracy. The experiment was carried out with four multi-objective test functions compared with two classical multi-objective algorithms, non-dominated sorting genetic algorithm and multi-objective particle swarm optimization algorithm. The results show that the proposed algorithm has excellent performance with faster convergence rate and strong ability to jump out of local optimum, could use to solve many optimization problems.

Particle Swarm Optimization Algorithm Based on Chaotic Sequences and Dynamic Self-Adaptive Strategy

To deal with the problems of premature convergence and tending to jump into the local optimum in the traditional particle swarm optimization, a novel improved particle swarm optimization algorithm was proposed. The self-adaptive inertia weight factor was used to accelerate the converging speed, and chaotic sequences were used to tune the acceleration coefficients for the balance between exploration and exploitation. The performance of the proposed algorithm was tested on four classical multi-objective optimization functions by comparing with the non-dominated sorting genetic algorithm and multi-objective particle swarm optimization algorithm. The results verified the effectiveness of the algorithm, which improved the premature convergence problem with faster convergence rate and strong ability to jump out of local optimum.

Metaheuristics for a bi-objective location-routing-problem in waste collection management

A waste collection problem is one of the most important problems in reverse logistics. This paper presents a mathematical model for waste collection problem and considers waste collection from customers’ location and waste disposal in compatible disposal facilities. The fleet of vehicles is heterogeneous and vehicles have multi-separated compartments. Vehicles have different capacity, different travel time and distance limitations, and different variable and fixed costs. Vehicles start their route from depots, collect wastes from customers’ location, and move to disposal centers to unload waste in the related disposal facility, and then return to depots. We should choose appropriate location for depots and disposal facilities from potential locations with respect to economic and social objectives. To solve the model, two well-known multi-objective evolutionary algorithms, namely non-dominated sorting genetic algorithm (NSGA-II) and multi-objective particle swarm optimization (MOPSO) are proposed. Finally, the computational results are compared and analyzed.

Optimal Design of Seismic Protective Devices for Highway Bridges Using Performance-Based Methodology and Multiobjective Genetic Optimization

AbstractThis study investigates the effectiveness and optimal design of protective devices for the seismic protection of highway bridges. The Painter Street Overcrossing is seismically redesigned with protective devices. Component-level fragility functions are first derived by probabilistic seismic demand analysis using nonlinear time history analyses that include soil–structure interaction effects and ground motion uncertainties. The bridge repair cost ratios are then derived using a performance-based methodology and the associated component failure probability. Results of the comparison of two initial protection designs show that the bridge repair cost ratios provided an efficient evaluation of the protective devices. Subsequently, a multiobjective genetic optimization method utilizing the Pareto optimal concept is employed to identify the optimal design parameters of protective devices for six design cases with various combinations of isolation bearings and fluid dampers. Finally, the repair cost ratio...

Hull mixed-model assembly line balancing using a multi-objective genetic algorithm simulated annealing optimization approach

Hull assembly line balancing has significant impact on performance of shipbuilding system and is usually a multi-objective optimization problem. In this article, the primary objectives of the hull assembly line balancing are to minimize the number of workstations, to minimize the static load balancing index, to minimize the dynamic load balancing index between workstations, and to minimize the multi-station-associated complexity. Because this problem comes under combinatorial optimization category and is non-deterministic polynomial-time hard, an improved genetic algorithm simulated annealing is presented. In genetic algorithm simulated annealing, the task sequence numbers are used as chromosomes, and selection, crossover, and mutation operators only deal with the elements of task set instead of the ones of the problem space. In order to prevent the algorithm appearing early convergence or getting local optimal result, the simulated annealing algorithm is used to deal with the individuals. Meanwhile, the algorithm is embedded with the hierarchical scheduling tactics in order to solve the selection problem on optimal solution in the Pareto-optimal set. A number of benchmark problems are solved to prove the superior efficiency of the proposed algorithm. Finally, a case study of the optimization of a hull assembly line was given to illustrate the feasibility and effectiveness of the method.

Trade-off among mechanical properties and energy consumption in multi-pass friction stir processing of Al7075 alloy employing neural network–based genetic optimization

Friction stir processing is a novel material processing technique. In this study, neural network–based genetic optimization is applied to optimize the process performance in terms of post-friction stir processing mechanical properties of Al7075 alloy and the energy cost. At first, the experimental data regarding the properties (i.e. elongation, tensile strength and hardness) and the consumed electrical energy are obtained by conducting tests varying two process parameters, namely, feed rate and spindle speed. Then, a numerical model making use of empirical data and artificial neural networks is developed, and multiobjective multivariable genetic optimization is applied to find a trade-off among the performance measures of friction stir processing. For this purpose, the properties like elongation, tensile strength and hardness are maximized and the cost of consumed electrical energy is minimized. Finally, the optimization results are verified by conducting experiments. It is concluded that artificial neural network together with genetic algorithm can be successfully employed to optimize the performance of friction stir processing.

Multi-objective genetic optimization of the heat transfer for tube inserted with porous media

This work is aimed at determining the best configurations of porous media inserted in a tube to achieve the excellent performance of fluid flow and heat transfer. Both geometry parameters and property parameters are the variables to be optimized. In order to evaluate the performance of tube, two conflicting objectives, Nusselt number Nu and friction factor f, are simultaneously considered. In the optimization process, computational fluid dynamics (CFD) and multi-objective genetic algorithm are coupled to obtain the numerical solutions of two-dimensional calculation model and Pareto front composed of non-inferior solutions. Besides, an attempt of utilizing limited resources is made by applying the penalty function instead of adding or modifying governing equations to restrain the flow resistance. Subsequently, technique for order preference by similarity to an ideal solution (TOPSIS) is employed to help decision makers determine the best alternative from the Pareto front. The results selected by TOPSIS are compared with the alternative which has the maximum Nusselt number. It is found that TOPSIS is effective to balance deferent objectives and determine compromise parameters.

Hybrid genetic multi objective/fuzzy algorithm for optimal sizing and allocation of renewable DG systems

This paper proposes a solution methodology for optimal allocation and sizing of renewable based DG units in radial distribution systems. The objectives are to minimize total power losses, to minimize system's investment cost and to improve voltage stability. The multi-objective genetic optimization algorithm is used to generate the optimal Pareto front and a fuzzy decision making function is used, as a hybrid function, to obtain the best compromise solution. The proposed technique is applied to typical IEEE bus systems including 33-bus system, 69-bus system and 119-bus system. This technique is implemented using different scenarios including all possible combinations of renewable DG units; to prove the effectiveness of the proposed methodology. The effect of DG based renewable energy allocation on the most sensitive buses to collapse is also carried out considering different load models. Copyright © 2016 John Wiley & Sons, Ltd.

Multiobjective Genetic Optimization Approach to Identify Pipe Segment Replacements and Inline Storages to Reduce Sanitary Sewer Overflows

Sanitary sewer overflows (SSOs) is the unintentional discharge of untreated sewage from the sanitary sewer system and pose serious risk to public health and to the environment. Rehabilitation plans to reduce SSOs involve increasing conveyance capacity and shaving peak flow using detention storages. Identifying the best location for rehabilitating the sanitary sewer network is a difficult task because of the great length of sanitary sewer systems. This study utilized single and multiobjective genetic algorithms (GAs) to design rehabilitation strategies for SSOs reduction in an existing sewer network. The Nondominated Sorting Genetic Algorithm II was linked to the EPA-SWMM to generate non-dominated sets of solutions that characterizes the tradeoffs between reduction in number of SSOs and cost (Case I), and the tradeoff between of volume of SSOs and cost (Case II). The results show that, when maximizing the reduction of number SSOs, the algorithm target first regions of the network with higher density of SSOs. When maximizing the reduction of volume of SSOs, the solutions prioritize the nodes with the largest overflow volumes. The tested approach provides a range of options to decision makers that seek to reduce or eliminate SSOs in an existing sanitary sewer system.