Fitness Landscape Research Articles

Factors Influencing Learners' Participation in Physical Education in The New Normal: An Input to Curriculum Development Program

The Philippines' current disaster mitigation and adaptation landscape this school year has brought a significant challenge to the education system as a result of the COVID-19 pandemic. Adversity and the resilience of people to hazards were experienced while combating this pandemic. Crisis and risk management systems are required to help prevent disease outbreaks and pandemics from affecting education communities. This point of view guides teachers, institutional heads, and authorities on how to deal with the crisis. The previous year saw many people face unexpected stress and adversity as a result of the worldwide outbreak of the coronavirus disease (COVID-19). As the number of challenges in providing basic education to students grows, school leaders and teachers must develop resilience in surviving and overcoming adversity. This study pursued to determine the relationship between school heads' managerial competencies to teachers' positive resilience in the new normal. A descriptive-correlational research design determined the relationship between the level of school heads' managerial competencies, and the level of teachers' positive resilience in the new normal within the four City Schools Divisions in Laguna for the school year 2021-2022. The study included 503 public school elementary teachers in the positions of teachers I-III and master teachers I-II as respondents. Using the Competency Framework for Southeast Asian School Heads (2014), and The Resilience Scale for Adults (2015), data gathering via an online platform was utilized. Frequency, percentage, mean, standard deviation, Pearson product-moment of correlation, and hierarchical linear regression were the statistical tools utilized to analyze and interpret the data gathered. Findings revealed that the school heads were practicing well in each construct of managerial competencies and each construct of positive resilience. The level of teachers' positive resilience in the new normal resulted in a very high level in terms of establishing friendly relationships easily. The researcher, therefore, concluded that when school heads demonstrated a high level of managerial competencies in terms of strategic thinking and innovation, instructional leadership, personal excellence, stakeholders' engagement, and managerial competencies, a better positive resilience may be expected to all the teachers in the new normal.

Hybridization and Transgressive Evolution Generate Diversity in an Adaptive Radiation of Anolis Lizards.

Interspecific hybridization may act as a major force contributing to the evolution of biodiversity. Although generally thought to reduce or constrain divergence between 2 species, hybridization can, paradoxically, promote divergence by increasing genetic variation or providing novel combinations of alleles that selection can act upon to move lineages toward new adaptive peaks. Hybridization may, then, play a key role in adaptive radiation by allowing lineages to diversify into new ecological space. Here, we test for signatures of historical hybridization in the Anolis lizards of Puerto Rico and evaluate 2 hypotheses for the role of hybridization in facilitating adaptive radiation-the hybrid swarm origins hypothesis and the syngameon hypothesis. Using whole genome sequences from all 10 species of Puerto Rican anoles, we calculated D and f-statistics (from ABBA-BABA tests) to test for introgression across the radiation and employed multispecies network coalescent methods to reconstruct phylogenetic networks that allow for hybridization. We then analyzed morphological data for these species to test for patterns consistent with transgressive evolution, a phenomenon in which the trait of a hybrid lineage is found outside of the range of its 2 parents. Our analyses uncovered strong evidence for introgression at multiple stages of the radiation, including support for an ancient hybrid origin of a clade comprising half of the extant Puerto Rican anole species. Moreover, we detected significant signals of transgressive evolution for 2 ecologically important traits, head length and toepad width, the latter of which has been described as a key innovation in Anolis. [Adaptive radiation; introgression; multispecies network coalescent; phenotypic evolution; phylogenetic network; reticulation; syngameon; transgressive segregation.].

Protein Fitness Prediction Is Impacted by the Interplay of Language Models, Ensemble Learning, and Sampling Methods.

Advances in machine learning (ML) and the availability of protein sequences via high-throughput sequencing techniques have transformed the ability to design novel diagnostic and therapeutic proteins. ML allows protein engineers to capture complex trends hidden within protein sequences that would otherwise be difficult to identify in the context of the immense and rugged protein fitness landscape. Despite this potential, there persists a need for guidance during the training and evaluation of ML methods over sequencing data. Two key challenges for training discriminative models and evaluating their performance include handling severely imbalanced datasets (e.g., few high-fitness proteins among an abundance of non-functional proteins) and selecting appropriate protein sequence representations (numerical encodings). Here, we present a framework for applying ML over assay-labeled datasets to elucidate the capacity of sampling techniques and protein encoding methods to improve binding affinity and thermal stability prediction tasks. For protein sequence representations, we incorporate two widely used methods (One-Hot encoding and physiochemical encoding) and two language-based methods (next-token prediction, UniRep; masked-token prediction, ESM). Elaboration on performance is provided over protein fitness, protein size, and sampling techniques. In addition, an ensemble of protein representation methods is generated to discover the contribution of distinct representations and improve the final prediction score. We then implement multiple criteria decision analysis (MCDA; TOPSIS with entropy weighting), using multiple metrics well-suited for imbalanced data, to ensure statistical rigor in ranking our methods. Within the context of these datasets, the synthetic minority oversampling technique (SMOTE) outperformed undersampling while encoding sequences with One-Hot, UniRep, and ESM representations. Moreover, ensemble learning increased the predictive performance of the affinity-based dataset by 4% compared to the best single-encoding candidate (F1-score = 97%), while ESM alone was rigorous enough in stability prediction (F1-score = 92%).

Protein–protein interactions in the Mla lipid transport system probed by computational structure prediction and deep mutational scanning

The outer membrane (OM) of Gram-negative bacteria is an asymmetric bilayer that protects the cell from external stressors, such as antibiotics. The Mla transport system is implicated in the Maintenance of OM Lipid Asymmetry by mediating retrograde phospholipid transport across the cell envelope. Mla uses a shuttle-like mechanism to move lipids between the MlaFEDB inner membrane complex and the MlaA-OmpF/C OM complex, via a periplasmic lipid-binding protein, MlaC. MlaC binds to MlaD and MlaA, but the underlying protein-protein interactions that facilitate lipid transfer are not well understood. Here, we take an unbiased deep mutational scanning approach to map the fitness landscape of MlaC from Escherichia coli, which provides insights into important functional sites. Combining this analysis with AlphaFold2 structure predictions and binding experiments, we map the MlaC-MlaA and MlaC-MlaD protein-protein interfaces. Our results suggest that the MlaD and MlaA binding surfaces on MlaC overlap to a large extent, leading to a model in which MlaC can only bind one of these proteins at a time. Low-resolution cryo-electron microscopy (cryo-EM) maps of MlaC bound to MlaFEDB suggest that at least two MlaC molecules can bind to MlaD at once, in a conformation consistent with AlphaFold2 predictions. These data lead us to a model for MlaC interaction with its binding partners and insights into lipid transfer steps that underlie phospholipid transport between the bacterial inner and OMs.

Plug & play directed evolution of proteins with gradient-based discrete MCMC

A long-standing goal of machine-learning-based protein engineering is to accelerate the discovery of novel mutations that improve the function of a known protein. We introduce a sampling framework for evolving proteins in silico that supports mixing and matching a variety of unsupervised models, such as protein language models, and supervised models that predict protein function from sequence. By composing these models, we aim to improve our ability to evaluate unseen mutations and constrain search to regions of sequence space likely to contain functional proteins. Our framework achieves this without any model fine-tuning or re-training by constructing a product of experts distribution directly in discrete protein space. Instead of resorting to brute force search or random sampling, which is typical of classic directed evolution, we introduce a fast Markov chain Monte Carlo sampler that uses gradients to propose promising mutations. We conduct in silico directed evolution experiments on wide fitness landscapes and across a range of different pre-trained unsupervised models, including a 650 M parameter protein language model. Our results demonstrate an ability to efficiently discover variants with high evolutionary likelihood as well as estimated activity multiple mutations away from a wild type protein, suggesting our sampler provides a practical and effective new paradigm for machine-learning-based protein engineering.

Double-replica theory for evolution of genotype-phenotype interrelationship

The relationship between genotype and phenotype plays a crucial role in determining the function and robustness of biological systems. Here the evolution progresses through the change in genotype, whereas the selection is based on the phenotype, and the genotype-phenotype relation also evolves. The theory for such phenotypic evolution remains poorly developed, in contrast to evolution under the fitness landscape determined by genotypes. Here we provide a statistical-physics formulation of this problem by introducing replicas for genotype and phenotype. We apply it to an evolution model in which phenotypes are given by spin configurations; genotypes are an interaction matrix for spins to give the Hamiltonian, and the fitness depends only on the configuration of a subset of spins called the target. We describe the interplay between the genetic variations and phenotypic variances by noise in this model by our approach that extends the replica theory for spin glasses to include a spin replica for phenotypes and a coupling replica for genotypes. Within this framework we obtain a phase diagram of the evolved phenotypes against the noise and selection pressure, where each phase is distinguished by the fitness and overlaps for genotypes and phenotypes. Among the phases, a robust fitted phase, relevant to biological evolution, is achieved under the intermediate level of noise (temperature), where robustnesses to noise and to genetic mutation are correlated, as a result of replica symmetry. We also find a tradeoff between maintaining a high fitness level of phenotype and acquiring a robust pattern of genes as well as the dependence of this tradeoff on the ratio between the size of the functional (target) part to that of the remaining nonfunctional (non-target) one. The selection pressure needed to achieve high fitness increases with the fraction of target spins.

Rare-event sampling analysis uncovers the fitness landscape of the genetic code.

The genetic code refers to a rule that maps 64 codons to 20 amino acids. Nearly all organisms, with few exceptions, share the same genetic code, the standard genetic code (SGC). While it remains unclear why this universal code has arisen and been maintained during evolution, it may have been preserved under selection pressure. Theoretical studies comparing the SGC and numerically created hypothetical random genetic codes have suggested that the SGC has been subject to strong selection pressure for being robust against translation errors. However, these prior studies have searched for random genetic codes in only a small subspace of the possible code space due to limitations in computation time. Thus, how the genetic code has evolved, and the characteristics of the genetic code fitness landscape, remain unclear. By applying multicanonical Monte Carlo, an efficient rare-event sampling method, we efficiently sampled random codes from a much broader random ensemble of genetic codes than in previous studies, estimating that only one out of every 1020 random codes is more robust than the SGC. This estimate is significantly smaller than the previous estimate, one in a million. We also characterized the fitness landscape of the genetic code that has four major fitness peaks, one of which includes the SGC. Furthermore, genetic algorithm analysis revealed that evolution under such a multi-peaked fitness landscape could be strongly biased toward a narrow peak, in an evolutionary path-dependent manner.

Fitting and evaluating univariate and multivariate models of within-lineage evolution.

The nature of phenotypic evolution within lineages is central to many unresolved questions in paleontology and evolutionary biology. Analyses of evolutionary time-series of ancestor-descendant populations in the fossil record are likely to make important contributions to many of these debates. However, the limited number of models that have been applied to these types of data may restrict our ability to interpret phenotypic evolution in the fossil record. Using uni- and multivariate models of trait evolution that make different assumptions regarding the dynamics of the adaptive landscape, I evaluate contrasting hypotheses to explain evolution of size in the radiolarian Eucyrtidium calvertense and armor in the stickleback Gaserosteus doryssus. Body size evolution in E. calvertense is best explained by a model where the lineage evolves as a consequence of a shift in the adaptive landscape that coincides with the initiation of neosympatry with its sister lineage. Multivariate evolution of armor traits in a stickleback lineage (Gasterosteus doryssus) show evidence of adaptation towards independent optima on the adaptive landscape at the same time as traits change in a correlated fashion. The fitted models are available in a the R package evoTS, which builds on the commonly used paleoTS framework.

Outcomes of multifarious selection on the evolution of visual signals.

Multifarious sources of selection shape visual signals and can produce phenotypic divergence. Theory predicts that variance in warning signals should be minimal due to purifying selection, yet polymorphism is abundant. While in some instances divergent signals can evolve into discrete morphs, continuously variable phenotypes are also encountered in natural populations. Notwithstanding, we currently have an incomplete understanding of how combinations of selection shape fitness landscapes, particularly those which produce polymorphism. We modelled how combinations of natural and sexual selection act on aposematic traits within a single population to gain insights into what combinations of selection favours the evolution and maintenance of phenotypic variation. With a rich foundation of studies on selection and phenotypic divergence, we reference the poison frog genus Oophaga to model signal evolution. Multifarious selection on aposematic traits created the topology of our model's fitness landscape by approximating different scenarios found in natural populations. Combined, the model produced all types of phenotypic variation found in frog populations, namely monomorphism, continuous variation and discrete polymorphism. Our results afford advances into how multifarious selection shapes phenotypic divergence, which, along with additional modelling enhancements, will allow us to further our understanding of visual signal evolution.

The foodplant fitness landscape of Hollow Rock Shelter, Western Cape, South Africa

Studying Stone Age foraging behaviours in terms of plant foods is difficult because of preservation, sampling and identification biases. Current foodplant populations, knowledge about their use in recent times, and how they are distributed across the landscape, provide valuable middle-range proxies from which work on archaeological landscapes and material can benefit. With this contribution we provide foodplant lists for three different foraging ranges (radii of ∼ 12.5 km, ∼35 km, and ∼ 70 km) around Hollow Rock Shelter. By comparing data for each of the foraging ranges directly, we discuss proportional increases in foodplant resources when moving further away from the site. We demonstrate that the ∼ 35 km foraging range is the most efficient. This implies that people staying at the site (for shorter or longer periods) may have regularly employed a strategy of temporary camping for a night or two away from the site to forage especially preservable foods that could be brought to the site. Our data highlight potential plant-food staples, and show that under-surface storage organs (USOs) of plants, followed by fruit and leaves are the most abundant edible plant parts available on the Hollow Rock Shelter landscape within all three of the foraging ranges, and that most of these could be eaten raw.

A fitness trade-off explains the early fate of yeast aneuploids with chromosome gains

The early development of aneuploidy from an accidental chromosome missegregation shows contrasting effects. On the one hand, it is associated with significant cellular stress and decreased fitness. On the other hand, it often carries a beneficial effect and provides a quick (but typically transient) solution to external stress. These apparently controversial trends emerge in several experimental contexts, particularly in the presence of duplicated chromosomes. However, we lack a mathematical evolutionary modeling framework that comprehensively captures these trends from the mutational dynamics and the trade-offs involved in the early stages of aneuploidy. Here, focusing on chromosome gains, we address this point by introducing a fitness model where a fitness cost of chromosome duplications is contrasted by a fitness advantage from the dosage of specific genes. The model successfully captures the experimentally measured probability of emergence of extra chromosomes in a laboratory evolution setup. Additionally, using phenotypic data collected in rich media, we explored the fitness landscape, finding evidence supporting the existence of a per-gene cost of extra chromosomes. Finally, we show that the substitution dynamics of our model, evaluated in the empirical fitness landscape, explains the relative abundance of duplicated chromosomes observed in yeast population genomics data. These findings lay a firm framework for the understanding of the establishment of newly duplicated chromosomes, providing testable quantitative predictions for future observations.

Abstract 1172: Oncogenic context shapes the fitness landscape of tumor suppression

Abstract Cancer progression is a quintessential example of a walk on an adaptive fitness landscape, with tumor growth depending on the cooperation of multiple driver mutations. While sequencing of tens of thousands of clinical samples has revealed a vast set of cancer drivers, far less is known about the interactions of oncogene-tumor suppressor pairs. Due to patient-level selection bias, confounding biological factors and limited sample sizes, a map of these interactions cannot be generated from human data alone and, instead, requires direct perturbational experiments and functional genomics approaches. Here, we model and quantify tumors using an autochthonous mouse platform and Tubaseq, which integrates barcoded lentiviral-sgRNA/Cre vectors and high-throughput barcode sequencing to uncover the number of neoplastic cells in each tumor of each genotype. Across four oncogenic contexts—KRAS G12D, KRAS G12C, BRAF V600E, and EGFR L858R—we analyzed >10,000,000 tumors to estimate the tumorigenesis potential of the oncogene alone as well as the effect of inactivating 28 tumor suppressor genes. We discovered that despite KRAS G12D producing >10x the tumors as G12C, tumor suppressor inactivations provide similar growth effects on both backgrounds. In contrast, although the intrinsic abilities of BRAF V600E and EGFR L858R to drive tumor development fall within the range of the KRAS variants, tumor suppressive effects are categorically different in the context of each oncogene. Many tumor suppressors show clear sign epistasis with the oncogenes, whereby inactivation is advantageous in one context and neutral or deleterious in another. Inactivation of some of the strongest tumor suppressors (e.g., Lkb1, Setd2, and Kmt2d) in KRAS-driven tumors strongly decreases tumor growth in the presence of oncogenic EGFR. While some of these epistatic effects are consistent with a textbook understanding of the RAS pathway, most cannot be predicted based on the linear oncogenic EGFR → KRAS → BRAF pathway model. Analyses of clinical genomics data from AACR Project GENIE confirm that high rates of passenger mutations in KRAS- and BRAF-driven lung tumors, among other factors, prevent the discovery of these interactions from human data alone. However, for EGFR-mutant lung cancers, which are less confounded by high mutational burden, the rates of coincident tumor suppressor mutation are highly correlated with tumor growth effects in our in vivo model. Thus, we find via causal experiments that the landscape of tumor suppression is highly dependent on oncogenic context, with a minority of tumor suppressive effects robust to changes in the oncogene. These findings suggest that the utility of a specific cancer mutation as a prognostic or predictive biomarker of patient outcomes will be dependent on coincident mutations in the tumor and highlight the utility of high-throughput, quantitative autochthonous mouse models in advancing our understanding of cancer biology. Citation Format: Lily M. Blair, Joseph M. Juan, Lafia Sebastian, Vy B. Tran, Wensheng Nie, Gregory D. Wall, Mehmet Gerceker, Ian K. Lai, Edwin A. Apilado, Gabriel Grenot, David Amar, Giorgia Foggetti, Mariana Do Carmo, Zeynep Ugur, Debbie Deng, Alex Chenchik, Maria Paz Zafra, Lukas E. Dow, Katerina Politi, Jonathan J. MacQuitty, Dmitri A. Petrov, Monte M. Winslow, Michael J. Rosen, Ian P. Winters. Oncogenic context shapes the fitness landscape of tumor suppression [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1172.

Epistasis decreases with increasing antibiotic pressure but not temperature.

Predicting mutational effects is essential for the control of antibiotic resistance (ABR). Predictions are difficult when there are strong genotype-by-environment (G × E), gene-by-gene (G × G or epistatic) or gene-by-gene-by-environment (G × G × E) interactions. We quantified G × G × E effects in Escherichia coli across environmental gradients. We created intergenic fitness landscapes using gene knock-outs and single-nucleotide ABR mutations previously identified to vary in the extent of G × E effects in our environments of interest. Then, we measured competitive fitness across a complete combinatorial set of temperature and antibiotic dosage gradients. In this way, we assessed the predictability of 15 fitness landscapes across 12 different but related environments. We found G × G interactions and rugged fitness landscapes in the absence of antibiotic, but as antibiotic concentration increased, the fitness effects of ABR genotypes quickly overshadowed those of gene knock-outs, and the landscapes became smoother. Our work reiterates that some single mutants, like those conferring resistance or susceptibility to antibiotics, have consistent effects across genetic backgrounds in stressful environments. Thus, although epistasis may reduce the predictability of evolution in benign environments, evolution may be more predictable in adverse environments. This article is part of the theme issue 'Interdisciplinary approaches to predicting evolutionary biology'.

Global epistasis on fitness landscapes.

Epistatic interactions between mutations add substantial complexity to adaptive landscapes and are often thought of as detrimental to our ability to predict evolution. Yet, patterns of global epistasis, in which the fitness effect of a mutation is well-predicted by the fitness of its genetic background, may actually be of help in our efforts to reconstruct fitness landscapes and infer adaptive trajectories. Microscopic interactions between mutations, or inherent nonlinearities in the fitness landscape, may cause global epistasis patterns to emerge. In this brief review, we provide a succinct overview of recent work about global epistasis, with an emphasis on building intuition about why it is often observed. To this end, we reconcile simple geometric reasoning with recent mathematical analyses, using these to explain why different mutations in an empirical landscape may exhibit different global epistasis patterns-ranging from diminishing to increasing returns. Finally, we highlight open questions and research directions. This article is part of the theme issue 'Interdisciplinary approaches to predicting evolutionary biology'.

Interdisciplinary approaches to predicting evolutionary biology.

Interdisciplinary approaches to predicting evolutionary biology.

Impact of population size on early adaptation in rugged fitness landscapes.

Owing to stochastic fluctuations arising from finite population size, known as genetic drift, the ability of a population to explore a rugged fitness landscape depends on its size. In the weak mutation regime, while the mean steady-state fitness increases with population size, we find that the height of the first fitness peak encountered when starting from a random genotype displays various behaviours versus population size, even among small and simple rugged landscapes. We show that the accessibility of the different fitness peaks is key to determining whether this height overall increases or decreases with population size. Furthermore, there is often a finite population size that maximizes the height of the first fitness peak encountered when starting from a random genotype. This holds across various classes of model rugged landscapes with sparse peaks, and in some experimental and experimentally inspired ones. Thus, early adaptation in rugged fitness landscapes can be more efficient and predictable for relatively small population sizes than in the large-size limit. This article is part of the theme issue 'Interdisciplinary approaches to predicting evolutionary biology'.

The Diepkloof Rock Shelter foodplant fitness landscape, Western Cape, South Africa

ABSTRACT This paper reports on, and supplies the checklists for, the known foodplant population currently growing around Diepkloof Rock Shelter, South Africa, within radii of ∼12.5 km, ∼35 km and ∼70 km. It demonstrates that many of the species identified from the shelter’s Middle Stone Age charcoal record still grow on the landscape today and that 33 of the ancient charcoal-identified species are historically known as foodplants in southern Africa. With this contribution the current foodplant species that could serve as dietary resources are extended to 81 within a day’s return trip (∼12.5 km) from the site, 158 within the ∼35 km radius and 170 on the greater foraging landscape extending up to ∼ 70 km from Diepkloof. Data are presented that indicate the proportional increases in plant growth forms and edible plant parts by extending foraging ranges. Initial interpretation of the data shows that foragers may have had adequate access to herbs and shrubs from which leafy vegetables and fruits can be collected, as well as hydrophytes (such as waterblommetjies) within a day’s return trip. To increase especially access to energy-rich corms and roots, they may have used additional/other camp sites or home bases to exploit their ∼35 km foraging range. By travelling over longer distances, access to grass grains that also serve as famine foods can be exponentially increased.

A Rolling-Adaptive Peak Clipping Control Strategy Coordinating RBE Recycling and PV Consumption

Electrified railway is one of the major users of electricity in the power system. The access of energy storage system (ESS) effectively solves the problems of peak traction load (TL) shock and regenerative braking energy (RBE) recycling, and the access of photovoltaic (PV) system improves environmental and economic benefits. However, in the current control strategy based on the electrified railway with ESS and PV, there are still problems of less optimal overall peak clipping effect of TL and insufficient penetration of renewable energy sources (RES) due to low utilization and lack of pre-processing capacity of ESS. Therefore, this paper proposes a rolling-adaptive peak clipping (RAPC) control strategy coordinating RBE recycling and PV consumption. Firstly, perform rolling peak clipping through dynamically divided implementation windows to make full use of the available energy in each time period. Secondly, according to the state of charge (SOC) of ESS and the specific working condition of TL, the ESS output is modified accordingly to perform adaptive deep peak clipping maximizing the consumption of PV energy and RBE. The simulation verified that the proposed strategy can improve the overall peak-clipping effect of TL, and achieve full recovery of RBE and complete consumption of PV energy under the same ESS configuration compared to the global optimization method. The results show that the RAPC control strategy can realize optimal peak-clipping and RBE utilization, and increase the PV penetration with limited ESS capacity, thus improving the economic benefits of the traction power supply system (TPSS) and promoting the energy transition of TPSS.

Reinforcement learning-based particle swarm optimization with neighborhood differential mutation strategy

The particle swarm optimization (PSO) algorithm has been one of the most effective methods for solving various engineering optimization problems. Most existing PSO variants frequently use fixed operators, the adoption of a fixed operator learning mode may restrict the intelligence level of each particle, thus reducing the performance of PSO in solving optimization issues with complicated fitness landscapes. To address single goal real-parameter numerical optimization while overcoming the above shortcoming, this paper proposes a reinforcement learning-based particle swarm optimization with neighborhood differential mutation strategy (NRLPSO). In NRLPSO, a dynamic oscillation inertial weight (DOW) strategy that provides particles with dynamic adjustment ability in different situations is designed. To resolve the operator selection conundrum of exploration and exploitation, a reinforcement learning-based velocity vector generation (VRL) strategy is developed. At each iteration, particles select the velocity update model based on reinforcement learning, and VRL helps to thoroughly search the problem space. A velocity updating mechanism based on cosine similarity (VCS) is applied to control the velocity learning mode to determine more promising solutions. Furthermore, to alleviate the problem of premature convergence, a local update strategy with neighborhood differential mutation (NDM) is employed to increase the diversity of the algorithm. To verify the efficiency of the proposed algorithm, the CEC2017 and CEC2022 test suites are implemented, and nine classic or state-of-the-art PSO variants are comprehensively tested. The experimental results show that NRLPSO outperforms the popular PSO variants in terms of convergence speed and accuracy. Since NRLPSO utilizes the DE mutations, it is compared with the representative LSHADE variant algorithm - LSHADE_SPACMA. Although LSHADE_SPACMA is better than NRLPSO concerning algorithm stability and convergence accuracy, we will refine our work in the future to enhance the performance in all aspects.

Unbalanced Magnetic Pull Disturbance Compensation of Magnetic Bearing Systems in MSCCs

Unbalanced magnetic pull (UMP) disturbance seriously affects the stable operation of the magnetically suspended rotor (MSR) system. This article proposes an UMP disturbance suppression method based on a linear extended state observer (LESO) and adaptive peak filter (APF). The cascade structure of LESO and APF can accurately estimate the UMP disturbance in the whole speed range. First, the dynamic model of the active magnetic bearing rotor systems is established and the UMP caused by the dynamic eccentricity in the MSR system is analyzed. Second, a third-order LESO and an APF are designed. The performance evaluation of the LESO indicates that estimation error of disturbance decreases with the increase of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\omega _{o}$</tex-math></inline-formula> , and the addition of APF can make UMP disturbance estimation unaffected by the high-frequency noise. After adding this method, the closed-loop system remains stable within the working rotation speed range. Finally, simulation results show that the UMP disturbance can be estimated and compensated. Experimental results validate the proposed method can well suppress the large vibration of the UMP on the rotor.