A Distributed Multiple Populations Framework for Evolutionary Algorithm in Solving Dynamic Optimization Problems

Abstract

Aiming to dynamic optimization problems (DOPs), this paper develops a novel general distributed multiple populations (DMP) framework for evolutionary algorithms (EAs). DMP employs six strategies designed in three levels (i.e., population-level, subpopulation-level, and individual-level) to deal with different kinds of DOPs. First, the population-level subpopulation division estimation strategy in initialization phase rationally divides the whole population into several subpopulations to explore distinct subareas of search space sufficiently. Then, during the steady evolutionary process, diversity preservation in individual-level and population-level accelerates the responsiveness of the whole population to a new landscape, while subpopulation-level self-learning of elitist individuals promotes the exploitation of promising areas. Moreover, in subpopulation-level, the archive quality assurance technique avoids repeat exploring the same peaks by storing the locations of different peaks with low redundancy. When landscape variation occurs, in population-level, historical information containing excellent evolutionary pattern is recorded to guide the population evolution better in the new environment. DMP framework is easy to implement in various EAs due to its well generality and independence about operators and parameters of the embedded algorithm. Four DMP-EAs are accomplished in this paper whose basic algorithms are particle swarm optimization (PSO) and differential evolution (DE) with different settings. The performance of the four proposed DMP-EAs is evaluated on all the widely used complex DOP benchmarks from CEC 2009. The testing results indicate that the DMP-EAs generally significantly outperform many state-of-the-art dynamic EAs (DEAs) on most of DOP benchmarks.INDEX TERMS Dynamic optimization problem (DOP), distributed multiple population (DMP) framework, multi-level diversity preservation, adaptive historical information utilization, dynamic evolutionary algorithm (DEA)

A survey on evolutionary computation for complex continuous optimization

“…AMP creates non-overlapping populations by a single linkage hierarchical clustering method and adopts an adaptive mechanism that learns from algorithm behavior changes through interacting with environments for dynamically adjusting the number of populations. Recently, Luo et al (2019) proposed a distributed multiple population framework to increase algorithm diversity for solving DOP. Zhang et al (2019b) proposed a new cluster-based clonal selection algorithm, where a max-min distance cluster method based on the fitness and Euclidean distance was used to partition the population.…”

A Survey of Evolutionary Continuous Dynamic Optimization Over Two Decades—Part A

“…where n a is the number of individuals in the population a, and c a is its center position, which is the average position of all members ( i∈a xi na ). This method has also been used in several DOAs, including [53]- [55]. In [56], the spatial size of a population is calculated similar to (P A -8), but the center position is replaced by the best found position in the population.…”

A Survey of Evolutionary Continuous Dynamic Optimization Over Two Decades—Part B

“…Best before change error [26], [27], [29], [41], [42], [51], [63], [72], [74], [75], [78], [82], [83], [87], [94], [97], [100], [106]- [108], [113]- [118], [121]- [123], [130], [134], [135], [137], [142]- [144], [147]- [162], [164]- [166], [168], [170], [172], [174], [194] Offline performance [18], [28], [35], [43], [62], [122], [126] Offline error (evaluation based) [15], [18], [22]- [25],…”