A Novel Multiobjective Optimization Method Based on Sensitivity Analysis

Abstract

For multiobjective optimization problems, different optimization variables have different influences on objectives, which implies that attention should be paid to the variables according to their sensitivity. However, previous optimization studies have not considered the variables sensitivity or conducted sensitivity analysis independent of optimization. In this paper, an integrated algorithm is proposed, which combines the optimization method SPEA (Strength Pareto Evolutionary Algorithm) with the sensitivity analysis method SRCC (Spearman Rank Correlation Coefficient). In the proposed algorithm, the optimization variables are worked as samples of sensitivity analysis, and the consequent sensitivity result is used to guide the optimization process by changing the evolutionary parameters. Three cases including a mathematical problem, an airship envelope optimization, and a truss topology optimization are used to demonstrate the computational efficiency of the integrated algorithm. The results showed that this algorithm is able to simultaneously achieve parameter sensitivity and a well-distributed Pareto optimal set, without increasing the computational time greatly in comparison with the SPEA method.

Application of Multi-objective Charged System Search Algorithm for Optimization Problems

“…i n j GG  xx (13) where ( ) max(0, ( )), n 1, 2,..., k nn Gg  xx (14) With Oyama's constraint-handling approach, was applied niching based on the number of constraint violations to infeasible solutions. Here, a standard fitness sharing [52] is applied to the infeasible designs based on their constraint violations as:…”

“…In single-objective optimization, the fitness of solutions is reachable easily due to existing just one objective function while for multi-objective optimization, no single unique solution can be determined as the best; instead, a set of non-dominated solutions should be found in order to get a good approximation of the true Pareto fronts [3,[12][13][14], which leads to trade-off among the objectives [15,16] .…”

“…3) Even if a multi-objective algorithm can find solutions on a Pareto front, there is no guarantee that distribution of multiple Pareto points becomes uniformly and this may reduce the applicability of the results [3,14,17].…”

“…After that, a wide variety of MOEAs have been suggested, such as Micro-Genetic Algorithm (Micro-GA) [19] , Non-dominated Sorting Genetic Algorithm (NSGA) [20] , New variant of NSGA or NSGA-II [21] , Strength Pareto Evolutionary Algorithm (SPEA) [22] , SPEA2 [23] , Pareto Archive Evolution Strategy (PAES) [24] , Pareto Differential Evolution Approach (PDEA) [25], MOEA/D: A Multiobjective Evolutionary Algorithm Based on Decomposition [26], NSGAII based on Differential Evolution (NSGAII-DE [27], A hybrid multi-objective particle swarm optimization and decision making procedure for optimal design of truss structures [28], The third version of Generalized Differential Evolution (GDE3) [29], Multi-Objective Differential 4 Evolution-the Ranking-based Mutation Operator (MODE-RMO) [30], Multi-Objective Particle Swarm Optimization (MOPSO) [31], Differential Evolution for Multi-Objective Optimization (DEMO) [32], A novel hybrid charge system search and particle swarm optimization method for multi-objective optimization [33], Multi-Objective Differential Evolution (MODE) [34], MultiObjective bees algorithms(Bees) [35] and Non-dominated Rank Genetic Algorithm (NRGA) [36]. Recent years, some other types of algorithms were also developed such as Multi-Objective Cuckoo Search (MOCS) [37], Multi-Objective Firefly Algorithm (MOFA) [38], A new multiswarm multi-objective optimization method for structural design [39], A swarm based memetic evolutionary algorithm for multi-objective optimization of large structures [40], Multi-Objective Flower Pollination Algorithms (MOFPA) [17] and Multi-objective Optimization Method Based on Sensitivity Analysis [14].…”

Calibrating a Process-based Model to Enhance Robustness in Carbon Sequestration Simulations: the Sase of Cedrus Atlantica (Endl.) Manetti ex Carrière

“…Sensitivity analysis defines the sensitivity as well as the importance of each model parameter and provides a sufficient basis for selection during model calibration. Optimizing parameters with low sensitivity increases the computation time without significantly improving the accuracy of the model [47].…”