Non-revisiting Genetic Algorithm Research Articles

Two-stage optimization method for design of reinforced concrete frames using optimal pre-determined section database and non-revisiting genetic algorithm

Two-stage optimization method for design of reinforced concrete frames using optimal pre-determined section database and non-revisiting genetic algorithm

Non-revisiting stochastic search revisited: Results, perspectives, and future directions

The non-revisiting genetic algorithm (NrGA) uses the entire search history with parameter-less adaptive mutation to significantly enhance search performance. In the past decade, a family of non-revisiting stochastic search (NrSS) methods has been developed. Using the entire (or partial) search history to assist evolutionary computation can achieve not only the goal of duplicated solution prevention, but also other utilizations such as adaptive parameter-less local search and fitness landscape estimation. In this survey, the focus is on the memory-assisted stochastic search techniques that store the search history in a binary space partitioning (BSP) tree. First, the basic NrGA is reviewed. Then, the development of the family of NrSS is reviewed from three aspects: 1) the basic NrSS algorithms, 2) conceptual extension of non-revisit and usage of the search history as novel operators and estimators, and 3) application of NrSS to different problem types, such as multi-objective problems, dynamic problems, and multi-modal problems. A comprehensive classification of search history-assisted algorithms suggests that the cumulative historical search information can be developed in a more functional manner; for example, a BSP tree can be simultaneously used for revisit prevention, fitness landscape estimation, and adaptive operation. Both the application on real-world problems and the theoretical analysis of NrSS are reviewed. Possible future work suggestions include a deeper understanding of the non-revisiting scheme in theory, a more comprehensive functional usage of search history, and a closer cooperation with data-driven techniques.

A non-revisiting genetic algorithm based on a novel binary space partition tree

As one of the most popular evolutionary algorithms for solving complex optimization problems, genetic algorithm has been extensively studied in the last three decades. Since genetic algorithm is a stochastic algorithm that may revisit duplicated solutions, it could suffer from low convergence speed on some real-world problems making algorithms likely to get trapped into local optimums. To address this issue, this paper proposes a non-revisiting genetic algorithm with a novel binary space partition (BSP) tree. The proposed BSP tree records all the generated solutions, which enables the algorithm to quickly determine whether a newly generated solution is duplicated or not. Moreover, the proposed algorithm fine-tunes the solutions according to the topology of the BSP tree in each generation, and thus can improve the population diversity and convergence speed. In comparison to six representative evolutionary algorithms, the proposed non-revisiting genetic algorithm exhibits better overall performance on eight benchmark problems, the power system fault diagnosis problem, and the molecular signatures selection problem.

Non-revisiting genetic algorithm with adaptive mutation using constant memory

The continuous non-revisiting genetic algorithm (cNrGA) uses the entire search history and parameter-less adaptive mutation to significantly enhance search performance. Storing the entire search history is natural and costs little when the number of fitness evaluations is small or moderate. However, if the number of evaluations required is substantial, some memory management is desirable. In this paper, we propose two pruning mechanisms to keep the memory used constant. They are least recently used pruning and random pruning. The basic idea is to prune a unit of memory when the memory threshold is reached and some new search information is required to be stored, thus keeping the overall memory used constant. Meanwhile, both pruning strategies naturally form parameter-less adaptive mutation operators. A study is carried out to evaluate the impact on performance caused by loss of search history information. Experimental results show that (1) both strategies can maintain the performance of cNrGA, up to the empirical limit when 90 % of the search history is not recorded, (2) cNrGA and its variants with constant memory outperform the real-coded genetic algorithm and the standard particle swarm optimization. By pre-extracting all the current prune-able history information and storing them into a list, namely, to-prune-list, the overhead of both pruning strategies becomes small. This suggests that cNrGA can be extended to use in situations when the number of fitness evaluations is much larger than before with no significant effect on statistical performance. This widens the applicability of cNrGA to include more practical problems that require larger number of fitness evaluations before converging to the global optima.

Determination of Power Distribution Network Configuration Using Non-Revisiting Genetic Algorithm

A non-revisiting genetic algorithm (NrGA) was used to determine distribution network configuration for loss reduction. By advocating binary space partitioning (BSP) to divide the search space and employing a novel BSP tree archive to store all the solutions that have been explored before, NrGA can quickly check for revisits by communicating with BSP tree archive when a new solution is generated by genetic algorithm (GA), and can mutate an alternative unvisited solution through a novel adaptive mutation mechanism that based on BSP tree while a revisit has occurred, which achieves no duplicates in the entire search. A method for getting independent loops of distribution network was realized using breadth-first search algorithm. Furthermore, the extended intermediate crossover mode, which requires no tuning parameter such as crossover rate and extends the crossover results, is employed for improving the performance of NrGA in solving distribution network configuration problem. The proposed approach has been successfully tested on three sample systems and three practical systems. Numerical studies have revealed its accuracy and efficient performance.

A Non-Revisiting Genetic Algorithm for Optimizing Numeric Multi-Dimensional Functions

A Non-Revisiting Genetic Algorithm for Optimizing Numeric Multi-Dimensional Functions