Papers by Keyword: Multi-Objective Optimization

Abstract: In building design, the need of optimization algorithms is considerably increasing due to the requirements of enhancing the overall performances, cost and sustainability objectives. An evolutionary algorithm coupled with building simulation code is often used. However, this approach is not widespread in actual application due to the high number of expensive simulation runs required by evolutionary algorithms. For this reason, the selection of an efficient and effective optimization algorithm becomes a key aspect in building design. In the literature there are several works analyzing the performance of different optimization algorithms, most of them by comparing the results obtained for the optimization of analytical test functions. However, there are no evidence-based studies deepening the efficiency and efficacy of these methods by comparing against the true solution of the discrete optimization problems on building design. This study compares the number of cost function evaluations and the percentage of the actual Pareto solutions of three algorithms used for the evaluation of the optimal refurbishment of three reference buildings for which the actual Pareto front is known through a brute-force approach.

Abstract: Integrated with orthogonal design method and numerical simulation, injection molding process of the Y-type electrical connectors was conducted to study the influence of process parameters on volume shrinkage rate and maximum warpage, which are regarded as product quality indices. The multi-indices valuation model for the main influencing factors of the process is developed. The influencing sensitivity to the multi-objective of the processing parameters, such as melt temperature, mold temperature, injection time and holding pressure, is determined by range analysis. Through analyzing the diagrams of influential factors, the optimized process parameter diagram is obtained and verified by simulation. The optimum parameters minimizing the warpage defect and shrinkage are: melt temperature (528K), mold temperature (338K), filling time (0.6s), holding pressure (100%) and holding time (10s). The results show that it is effective to balance the impact of process parameters on the shrinkage and warpage. The work can provide optimal design and process reference for the quality control and assembly precision.

Abstract: Fiber reinforced laminate design is a challenging problem in the field of composite laminates. It provides us a systematic way to design the laminates of desired properties while conveniently incorporating the thick-ness and mass constraints. In this paper, we pursue the multivariate graphite fiber reinforced laminate design problem using Ant Colony Optimization (ACO) algorithm. Classical lamination theory is used to determine mid-plane strains, curvatures and stresses in individual lamina under applied biaxial loading conditions. The fiber orientations, lamina thickness, number of layers and fiber volume fractions of lamina are considered as the optimization variables. Failure of the lamina is analyzed by Tsai–Wu failure criterion. Objective of the study is to maximize the load carry capacity of the composite laminate structure and minimize the areal mass density under multivariate/multiobjective optimization.

Abstract: In this paper, we address the mixed H₂/H_∞ control approach for linear time-invariant system based on linear matrix inequality (LMI). First, the problem to be solved is stated, and the satellite attitude dynamics is established and converted into a corresponding state space form. Then, the mixed H₂/H_∞ controller based on LMIs is designed in order to attain the state feedback gain matrix. To validate the efficiency and practicability of the proposed controller, simulation results based on satellite attitude system are presented, from which we can observe that under the condition of external disturbances, the system will be stable within 150s, and the maximum of control torque will be no more than 0.025Nm. Expanding the controller gain will affect the stabilizing process, but not the stabilization time, and it will increase the control input which will bring pressure to the actuator.

Abstract: This paper deals with a reconstruction of random media via multi-objective optimization. Two statistical descriptors, namely a two-point probability function and a two-point lineal path function, are repetitively evaluated for the original medium and the reconstructed image to appreciate the improvement in the optimization progress. Because of doubts of the weights setting in the weighted-sum method, purely multi-objective optimization routine Non-dominated Sorting Genetic Algorithm~II is utilized. Three operators are compared for creating new offspring populations that satisfy a prescribed volume fraction constraint. The main contribution is in the testing of the proposed methodology on several benchmark images.

Abstract: Multi-objective optimization of laser cutting for simultaneous improvement of performance characteristics is of great practical importance. In this study a range of value (ROV)-based Taguchi methodology is proposed for multi-objective optimization of laser cutting, i.e. surface roughness, kerf width and burr height in CO₂ laser cutting of AISI 304 stainless steel. Laser cutting experiment was conducted based on Taguchi’s L₂₇ experimental design by varying the laser power, cutting speed, assist gas pressure and focus position at three levels. In the proposed methodology based on the experimental data signal to noise ratios as per Taguchi’s method were calculated for each experimental trial upon which decision matrix was defined. Subsequently, multi-criteria decision making problem was solved by the ROV method. The proposed ROV-based Taguchi methodology has relatively simple computational procedure and can be easily applied by engineers for solving different multi-objective optimization problems that occur in real manufacturing environment.

Abstract: Non-conventional manufacturing techniques are most widely used in industries in order to achieve high accuracy and desirable product quality. Therefore, the selection of an appropriate machining parameter has become a crucial job before starting the operation. Several optimization methods are available to resolve the upstairs situation. The current study explores a novel technique namely multi-objective optimization on the basis of ratio analysis (MOORA) to solve different multi-objective problems that are encountered in the real-time manufacturing industries. This study focuses on the application of MOORA method for solving some non-conventional machining processes that have multiple criteria problems. Wire-Electric Discharge Machining (WEDM), Plasma Arc Cutting (PAC), Electro Chemical Micro Machining (ECMM), Electro Chemical Machining (ECM), Abrasive Jet Machining (AJM), Abrasive Water Jet Machining (AWJM), Ultrasonic Machining (USM), Laser Beam Machining (LBM) and Laser cutting process are the major attentions in the current study. Total nine NTM multi-criteria problems which include selection of proper machining parameters have been studied. The optimal settings of input variables obtained by using MOORA method nearly tie with those derived by the earlier investigators.

Abstract: The design of gears is critical for smooth running of any mechanism, automobile and machinery. Gear drive design starts with the need of optimizing the gear thickness, module, number of teeth etc., this creates huge challenges to a designer. Optimization algorithms are more flexible and gaining importance in engineering design problems, because of the accessibility and affordability of today’s mechanical field. A population based heuristic algorithm offers well-organized ways of creating and comparing a novel design solution in order to complete an optimal design. In this paper, a new artificial immune system based algorithm proposed as Modified Artificial Immune System (MAIS) algorithm is used to optimize a gear design problem. The results are compared with an existing design.

Abstract: According to the coordination and complementation of electric vehicles (EVs) and renewable resources, such as photovoltaic (PV) power generation, a micro gird system including EVs charging station and PV power generation was constructed firstly. Based on the target of maximizing the utilization ratio of PV power, considering the total cost of EVs charging, the time-of-use (TOU) price was introduced to establish the dual-objective optimization scheduling model of EVs charging. Furthermore, NSGA-II multi-objective optimization algorithm was applied to solve the model and the Pareto front of the non-dominated solutions was obtained. Finally, the optimized scheduling control strategy for EVs charging was proposed through normalized sorting of the non-dominated solutions. The optimal scheduling strategy could increase the utilization ratio of PV power on the basis of reducing the cost of EVs charging, promoting the local consumption of PV power.

Abstract: Job Shop scheduling should satisfy the constraints of time, order and resource. To solve this NP-Hard problem, multi-optimization for job shop scheduling problem (JSSP) in discrete manufacturing plant is researched. Objective of JSSP in discrete manufacturing enterprise was analyzed, and production scheduling optimization model was constructed with the optimization goal of minimizing the bottleneck machines’ make-span and the total products’ tardiness; Then, Particle Swarm Optimization (PSO) algorithm was used to solve this model by the process-based encoding mode; To solve the premature convergence problem of PSO, advantages of Simulated Annealing (SA) algorithm, such as better global optimization performance, was integrated into PSO algorithm and a Hybrid PSO-SA Algorithm (HPSA) was proposed and the flowchart was presented; Then, this hybrid algorithm was applied in actual production scheduling of a discrete manufacturing enterprise. Finally, comparative analysis of HPSA/SA/PSO optimal methods and actual scheduling plan was carried out, which verify the result that the HPSA is effective and superiority.