Solving Large-Scale Multiobjective Optimization Problems With Sparse Optimal Solutions via Unsupervised Neural Networks

Abstract

Due to the curse of dimensionality of search space, it is extremely difficult for evolutionary algorithms to approximate the optimal solutions of large-scale multiobjective optimization problems (LMOPs) by using a limited budget of evaluations. If the Pareto optimal subspace is approximated during the evolutionary process, the search space can be reduced and the difficulty encountered by evolutionary algorithms can be highly alleviated. Following the above idea, this paper proposes an evolutionary algorithm to solve sparse LMOPs by learning the Pareto optimal subspace. The proposed algorithm uses two unsupervised neural networks, a restricted Boltzmann machine and a denoising autoencoder to learn a sparse distribution and a compact representation of the decision variables, where the combination of the learnt sparse distribution and compact representation is regarded as an approximation of the Pareto optimal subspace. The genetic operators are conducted in the learnt subspace, and the resultant offspring solutions then can be mapped back to the original search space by the two neural networks. According to the experimental results on eight benchmark problems and eight real-world problems, the proposed algorithm can effectively solve sparse LMOPs with 10,000 decision variables by only 100,000 evaluations.

A Multiobjective Evolutionary Approach for Solving Large-Scale Network Reconstruction Problems via Logistic Principal Component Analysis

“…MAST-Net [24] is a memetic algorithm specially designed for NRPs, which has shown superiority on small-scale networks. SparseEA [63] and MOEA/PSL [60] have been validated to be effective on large-scale MOPs, and are implemented on the evolutionary multiobjective optimization platform PlatEMO [64]. For the comparative methods, we use identical parameters suggested in the original papers and codes.…”

Biobjective Optimization-Based Frequency Regulation of Power Grids with High-Participated Renewable Energy and Energy Storage Systems

“…en, the results can be sorted by adaptation value. When the number of noninferior solutions is less than the number of archive size N, copy the first N − |P t+1 | individuals X j with F(X j ) ≥ 1 from the resulting ordered list to P t+1 [11], and | • | corresponds to the number of elements of a set. (4) Archive truncation procedure: when the number of noninferior solutions exceeds the number of archive…”

“…To solve the BOCC with high nonlinearity, the metaheuristic-based multiobjective optimization algorithms show stronger global searching ability than the traditional mathematical optimization methods [10][11][12][13][14][15]. Widely used multiobjective algorithms include the improved strength Pareto evolutionary algorithm (SPEA2) [16] and a nondominated sorting genetic algorithm II (NSGA-II) [17].…”

A responsive ant colony optimization for large-scale dynamic vehicle routing problems via pheromone diversity enhancement

“…Evolutionary algorithms are among the most popular metaheuristics and have been demonstrated to be effective in various large-scale optimization [38,39,46]. Hence, we develop RACO under the framework of the evolutionary algorithm named ant colony optimization (ACO), which has also exhibited effectiveness in solving DVRPs as suggested in [29].…”