Self-adaptive differential evolution algorithm with improved mutation strategy

Wang, Shihao; Li, Yuzhen; Yang, Hongyu; Liu, Hong

doi:10.1007/s00500-017-2588-5

Cited by 60 publications

(28 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Step f14 Himmelblau In order to further validate the advantages of ELAPO, the ELAPO algorithm was compared with the basic LAPO algorithm and the other four algorithms that have better optimization results in recent years, including: All-dimension neighborhood based particle swarm optimization with randomly selected neighbors (ADN-RSN-PSO) [29], enhanced artificial bee colony algorithms with adaptive differential operators (ABCADE) [30], an optimization algorithm for teaching and learning based on hybrid learning strategies (DSTLBO) [31], and a self-adaptive differential evolution algorithm with improved mutation strategy (IMSaDE) [32].…”

Section: Analysis Of the Simulation Resultsmentioning

confidence: 99%

An Enhanced Lightning Attachment Procedure Optimization Algorithm

Wang

Jiang

2019

Algorithms

View full text Add to dashboard Cite

To overcome the shortcomings of the lightning attachment procedure optimization (LAPO) algorithm, such as premature convergence and slow convergence speed, an enhanced lightning attachment procedure optimization (ELAPO) algorithm was proposed in this paper. In the downward leader movement, the idea of differential evolution was introduced to speed up population convergence; in the upward leader movement, by superimposing vectors pointing to the average individual, the individual updating mode was modified to change the direction of individual evolution, avoid falling into local optimum, and carry out a more fine local information search; in the performance enhancement stage, opposition-based learning (OBL) was used to replace the worst individuals, improve the convergence rate of population, and increase the global exploration capability. Finally, 16 typical benchmark functions in CEC2005 are used to carry out simulation experiments with LAPO algorithm, four improved algorithms, and ELAPO. Experimental results showed that ELAPO obtained the better convergence velocity and optimization accuracy.

show abstract

Section: Analysis Of the Simulation Resultsmentioning

confidence: 99%

An Enhanced Lightning Attachment Procedure Optimization Algorithm

Wang

Jiang

2019

Algorithms

View full text Add to dashboard Cite

show abstract

“…In order to verify the performance of the algorithm, NMSIDE is used to optimize the 11 standard test functions [14,15,16] and compared with several recent typical DE algorithms. The 11 standard test functions are shown in table 1.…”

Section: Experimental Results and Analysismentioning

confidence: 99%

An Improved Differential Evolution Algorithm with Novel Mutation Strategy

Shen¹,

Zou²,

Zhang

2017

2017 2nd International Conference on Mechatronics and Information Technology (ICMIT 2017)

View full text Add to dashboard Cite

Aiming at the defects of differential evolution, such as premature convergence, low accuracy and other shortcomings, an improved differential evolution algorithm with novel mutation strategy(NMSIDE) is presented. The novel mutation strategy is used to avoid being trapped into local minima for NMSIDE. If the evolution of the individual stagnates, the individual will rely on the current best individual to move closer to the global best individual. The mutation rate varies dynamically within the range of values, and the crossover rate is dynamically varied based on the number of iterations. NMSIDE is tested on 11 standard functions and compared with the other state-of-the-art algorithms. The experimental results show that NMSIDE has higher convergence precision, faster convergence speed and better robustness.

show abstract

“…The performance of the DE algorithm depends largely on the proper choice of the scale factor, which can affect the convergence. In this context, a higher value of the scale factor, in the early stage, improves the exploration ability, which has a positive effect on the population diversity and, thus, avoids premature convergence at local optima [18,39]. In contrast, a smaller value of the scale factor, in the later stage, improves the exploitation ability, which can enhance the convergence speed [18,39].…”

Section: Differential Evolution Algorithm and Thementioning

confidence: 99%

Passive Target Localization Problem Based on Improved Hybrid Adaptive Differential Evolution and Nelder-Mead Algorithm

2020

View full text Add to dashboard Cite

This paper considers a passive target localization problem in Wireless Sensor Networks (WSNs) using the noisy time of arrival (TOA) measurements, obtained from multiple receivers and a single transmitter. The objective function is formulated as a maximum likelihood (ML) estimation problem under the Gaussian noise assumption. Consequently, the objective function of the ML estimator is a highly nonlinear and nonconvex function, where conventional optimization methods are not suitable for this type of problem. Hence, an improved algorithm based on the hybridization of an adaptive differential evolution (ADE) and Nelder-Mead (NM) algorithms, named HADENM, is proposed to find the estimated position of a passive target. In this paper, the control parameters of the ADE algorithm are adaptively updated during the evolution process. In addition, an adaptive adjustment parameter is designed to provide a balance between the global exploration and the local exploitation abilities. Furthermore, the exploitation is strengthened using the NM method by improving the accuracy of the best solution obtained from the ADE algorithm. Statistical analysis has been conducted, to evaluate the benefits of the proposed modifications on the optimization performance of the HADENM algorithm. The comparison results between HADENM algorithm and its versions indicate that the modifications proposed in this paper can improve the overall optimization performance. Furthermore, the simulation shows that the proposed HADENM algorithm can attain the Cramer-Rao lower bound (CRLB) and outperforms the constrained weighted least squares (CWLS) and differential evolution (DE) algorithms. The obtained results demonstrate the high accuracy and robustness of the proposed algorithm for solving the passive target localization problem for a wide range of measurement noise levels.

show abstract

Self-adaptive differential evolution algorithm with improved mutation strategy

Cited by 60 publications

References 25 publications

An Enhanced Lightning Attachment Procedure Optimization Algorithm

An Enhanced Lightning Attachment Procedure Optimization Algorithm

An Improved Differential Evolution Algorithm with Novel Mutation Strategy

Passive Target Localization Problem Based on Improved Hybrid Adaptive Differential Evolution and Nelder-Mead Algorithm

Contact Info

Product

Resources

About