Hysteresis compensation and adaptive control based evolutionary neural networks for piezoelectric actuator

As an effective global optimization technique, artificial bee colony (ABC) algorithm has become one of the hottest research topics in the fields of evolutionary algorithms. However, the solution search equation is not rotationally invariant, which causes the problem that the performance of ABC is sensitive to the coordinate system. Although many improved ABC variants have been developed, they rarely considered the problem. Hence, to solve the problem, we propose a new ABC variant with bi‐coordinate systems (BSABC), including the original coordinate system and the eigen coordinate system. The two coordinate systems own different characteristics: (1) the former one aims to maintain the population diversity, and (2) the latter one is to adapt the search to the fitness landscape of the problems. Based on the characteristics, in the BSABC, the two coordinate systems are used in the employed bee phase and onlooker bee phase, respectively. Meanwhile, two new solution search equations are designed by utilizing the elite information, and they are respectively performed in the two coordinate systems to further improve the algorithm performance. As another contribution of this study, in the scout bee phase, the multivariate Gaussian distribution is constructed to replace the original method to generate offspring, which is helpful to save the search experience. The performance of the BSABC is verified by extensive experiments on the CEC2013 test suite and one real‐world optimization problem, and four well‐established ABC variants and three other evolutionary algorithms are included in the performance comparison. The comparison results confirm that the BSABC shows competitive performance by achieving better results on the majority of test functions.

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Artificial bee colony algorithm with bi‐coordinate systems for global numerical optimization

Zhou

Huang

Tang

et al. 2022

“…For such problems, the optimal solution can be obtained using an accurate algorithm. 1,2 When the size of the solution space of optimization problems gradually increases, it will become very difficult or even impossible to solve these problems using accurate algorithms. Therefore, metaheuristic algorithm came into being.…”

Section: Introductionmentioning

confidence: 99%

“…For structured optimization problems, the scale of the solution space can be controlled. For such problems, the optimal solution can be obtained using an accurate algorithm 1,2 . When the size of the solution space of optimization problems gradually increases, it will become very difficult or even impossible to solve these problems using accurate algorithms.…”

Section: Introductionmentioning

confidence: 99%

Adaptive complex network topology with fitness distance correlation framework for particle swarm optimization

Sun

Huang

et al. 2021

The particle swarm optimization algorithm is an effective tool to solve various optimization problems due to the small number of parameters and the simple learning strategy. However, the updated strategy from the basic PSO mainly aims to learn the global optimal particles, and it often leads to premature convergence with poor solution accuracy. An adaptive complex network topology with a fitness distance correlation for the particle swarm optimization algorithm is proposed (CNAPSO). Using the CNAPSO algorithm, it is concluded that different network topologies have different degrees of dispersion in the process of particle swarm optimization search. Therefore, the adaptive strategy with the fitness distance correlation proposes to effectively balance the global exploration and local exploitation capabilities, which is the particle swarm adaptive network neighborhood topology. The neighborhood topology construction strategy with a complex network is used to construct the neighborhood topology for each particle. Therefore, the local optimal particles in the neighborhood participate in the search process of particle swarm optimization and eliminate the situation of only learning the global optimal particles. Moreover, it improves the solution accuracy of the particle swarm optimization algorithm. In addition, to avoid the particle swarm falling into premature

“…As a result, clients are released from being anchored on a fixed training set. The advantages of

sans-serifOGA

are as follows: (1) Universality : Our method is applicable over most neural network constructions, 22–25 including the most popular Convolutional Neural Network (CNN) and long short‐term memory (LSTM) models. Besides, our method does not make any requirements on the ED's local search algorithm.…”

Section: Introductionmentioning

confidence: 99%

Secure and efficient parameters aggregation protocol for federated incremental learning and its applications

Wang

Liang

Koe

et al. 2021

Federated Learning (FL) enables the deployment of distributed machine learning models over the cloud and Edge Devices (EDs) while preserving the privacy of sensitive local data, such as electronic health records. However, despite FL advantages regarding security and flexibility, current constructions still suffer from some limitations. Namely, heavy computation overhead on limited resources EDs, communication overhead in uploading converged local models' parameters to a centralized server for parameters aggregation, and lack of guaranteeing the acquired knowledge preservation in the face of incremental learning over new local data sets. This paper introduces a secure and resourcefriendly protocol for parameters aggregation in federated incremental learning and its applications. In this study, the central server relies on a new method for parameters aggregation called orthogonal gradient aggregation. Such a method assumes constant changes of each local data set and allows updating parameters in the orthogonal direction of previous parameters spaces.As a result, our new construction is robust against catastrophic forgetting, maintains the federated neural network accuracy, and is efficient in computation and