BiomimeticsPub Date : 2026-10-16DOI: 10.3390/biomimetics11100748
Sheng Liu, Huijuan Wang, Yijuan Yuan, Min Zhou, Yonghong Deng
{"title":"Adaptive Reinforced Elite Gray Wolf Optimization for Robotic Path Planning and Football Pass-Quality Prediction.","authors":"Sheng Liu, Huijuan Wang, Yijuan Yuan, Min Zhou, Yonghong Deng","doi":"10.3390/biomimetics11100748","DOIUrl":"10.3390/biomimetics11100748","url":null,"abstract":"<p><p>Complex optimization problems in engineering and data-driven modeling require effective exploration, stable convergence, and adaptability across different decision representations. This study proposes an Adaptive Reinforced Elite Gray Wolf Optimizer (AREGWO) that combines opposition-based initialization, nonlinear convergence control, reinforced elite guidance, adaptive weak perturbation, stagnation-aware partial restart, and greedy elite inheritance. AREGWO is evaluated through CEC2022 benchmark optimization, synthetic 2D constrained robotic path planning, and football pass-quality prediction. On CEC2022, AREGWO ranks second among nine algorithms in the Friedman test and significantly outperforms seven of eight competitors in the Wilcoxon signed-rank test, while L-SHADE remains the strongest baseline. Component-wise ablation shows that the six mechanisms provide complementary and problem-dependent contributions. In robotic path planning, 30 independent runs on six random maps yield a six-map average path length of 134.38 m, which is 3.61-5.12% shorter than those obtained by the four population-based baselines and 1.08% shorter than A*, with a 100% success rate in all tested environments. For football pass-quality prediction, AREGWO is coupled with TabM for outer-level hyperparameter optimization of a predefined pass-quality reference score. AREGWO-TabM achieves an RMSE of 0.445, an MSE of 0.198, and an R<sup>2</sup> of 0.920, providing the best RMSE, MSE, and R<sup>2</sup> among the evaluated neural models, although ResNet attains a lower MAE. These results demonstrate the competitive search capability of AREGWO across the numerical, synthetic spatial, and model-level optimization settings evaluated in this study.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702155/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261743","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-16DOI: 10.3390/biomimetics11100745
Zhiyuan Hu, Fulin Du, Dingsheng Feng, Qiuxuan Wu, Xiaoqing Tian, Jian Wang
{"title":"Design and Experimental Validation of a Biomimetic Jellyfish Robot with Asymmetric-Stiffness Swimming Arms and Dual-Power Cooperative Drive.","authors":"Zhiyuan Hu, Fulin Du, Dingsheng Feng, Qiuxuan Wu, Xiaoqing Tian, Jian Wang","doi":"10.3390/biomimetics11100745","DOIUrl":"10.3390/biomimetics11100745","url":null,"abstract":"<p><p>This study presents a biomimetic jellyfish robot integrating asymmetric-stiffness swimming arms with motor-spring cooperative actuation. Motor-driven spring charging and rapid spring release organize the recovery, power-stroke, and gliding phases. Hinged arm segments, return micro-springs, and a limiting structure are designed to permit recovery-stroke bending while resisting bending during the power stroke. Component CFD and MATLAB(R2024a) motion-cycle analysis provide preliminary hydrodynamic and structural configuration comparisons. An arithmetic check of the retained drag fits gives a velocity-dependent arm-drag reduction of approximately 52.87-54.41% over 0.10-0.30 m/s and a separate body-drag reduction of approximately 44.49%; these are fitted-model comparisons, not measured drag reductions or mesh-independent predictions. In the recorded water-tank experiments, the prototype traveled approximately 1.8 m over approximately 9 s, corresponding to an overall average straight-line speed of approximately 20 cm/s. The average speed during the combined power-stroke and gliding phases was approximately 27 cm/s, and the maximum turning angular velocity was approximately 4.2°/s. These observations support the engineering feasibility of the integrated prototype under the reported test conditions. Quantitative stiffness characterization, independent trial variability, and matched-timing model validation are not established from the retained records.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702345/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261882","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-15DOI: 10.3390/biomimetics11100743
Xuan Liu, Yuqing Yang, Tong Zhao
{"title":"Identifying Anticancer Peptide via a Semi-Supervised Prediction Framework with Bio-Inspired Augmentation.","authors":"Xuan Liu, Yuqing Yang, Tong Zhao","doi":"10.3390/biomimetics11100743","DOIUrl":"10.3390/biomimetics11100743","url":null,"abstract":"<p><p>Anticancer peptides (ACPs) have emerged as a promising class of therapeutic molecules due to their low toxicity and high specificity. However, experimentally identifying ACPs is labor-intensive and low-throughput, making virtual screening with intelligent models an attractive alternative. Despite recent progress, the number of ACP samples available in existing databases remains limited, which may cause predictive models to overfit the labeled information and exhibit suboptimal generalization. To address this issue, we propose a semi-supervised prediction framework with bio-inspired data augmentation for ACP identification, called AugACP. Specifically, we incorporate semantic knowledge into masked Transformer pretraining to guide ACP sequence augmentation and provide more reliable supervised signals. Based on the augmented data, we adopt a semi-supervised learning strategy that selects high-confidence pseudo-labels from augmented peptides to improve prediction performance. In the computational experiments, AugACP outperforms state-of-the-art baseline methods under various experimental configurations, particularly in generalizing to unseen oligopeptides. Moreover, t-SNE visualizations indicate that AugACP learns more discriminative embeddings for separating ACPs from non-ACPs, while external validation on novel ACPs also achieves an F1-score of 0.9272. These results highlight the predictive power of AugACP and its potential value in promoting anticancer drug discovery.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702777/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261861","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-15DOI: 10.3390/biomimetics11100740
Anghel Constantin, Roxana-Mariana Nechita, Oliver Ulerich, Dimitrie Constantin, Dana-Corina Deselnicu, Mirona Ana Maria Ichimov
{"title":"Spatial Decoupling of Foot-Substrate Cross-Talk in Plantar Pressure Sensor Arrays.","authors":"Anghel Constantin, Roxana-Mariana Nechita, Oliver Ulerich, Dimitrie Constantin, Dana-Corina Deselnicu, Mirona Ana Maria Ichimov","doi":"10.3390/biomimetics11100740","DOIUrl":"10.3390/biomimetics11100740","url":null,"abstract":"<p><p>Measuring plantar pressure with smart insoles is widely used in gait analysis, medical rehabilitation, and clinical monitoring, but the continuous elastomeric layers of hybrid footwear transmit force between adjacent sensors and produce mechanical cross-talk that distorts local readings. The objective of this study was to develop a deterministic framework that identifies the spatial mechanical influence between sensors and compensates for it after signal acquisition. Controlled vertical loads were applied individually to each sensor, while the responses of all sensors were recorded each time as a vector and placed in a matrix; the resulting force data, complemented by a distance-dependent transfer function, formed the direct influence matrix of the Decision-Making Trial and Evaluation Laboratory (DEMATEL) method. Normalization and inversion generated a static decoupling operator applied to the vector of raw measurements. Secondary signals, which initially ranged from 40 N to 970 N, were reduced to a residual band from 3 N to 75 N (97.8% attenuation). The system, once calibrated in this way, provides a clearer spatial representation of local plantar pressure without applying changes to the physical structure of the insole and can be executed through a single matrix-vector multiplication on microcontrollers used in wearable platforms with limited resources.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702658/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-15DOI: 10.3390/biomimetics11100742
Yuyang Wei, Weijie Fei, Jiarong Wang, Luzheng Bi
{"title":"Bioreactor Optimisation Through Coupled Fluid-Structure Interaction and Data-Driven Analysis: A Biomimetic Framework for Predicting Early Cell Attachment.","authors":"Yuyang Wei, Weijie Fei, Jiarong Wang, Luzheng Bi","doi":"10.3390/biomimetics11100742","DOIUrl":"10.3390/biomimetics11100742","url":null,"abstract":"<p><p>Early cell attachment sets the baseline for proliferation, matrix deposition and construct maturation in tissue engineering, yet conventional bioreactors deliver simplified, quasi-static stimuli that poorly represent the multi-axial mechanical environment of native tissue. Here we reported a biomimetic computational-experimental framework for a humanoid-robot-driven perfusion bioreactor. The framework coupled a high-fidelity fluid-structure interaction (FSI) model of the culture chamber with a machine-learning model in order to describe the initial 3 h attachment phase, which preceded the application of robotic multi-axial loading. All experimental data reported here were therefore acquired during the static and perfused seeding phase; robotic loading was applied subsequently and was not part of the present dataset. Local mechanical and fluidic fields-wall shear stress, flow velocity, pressure, von Mises stress and scaffold strain-were mapped across ten scaffold subregions and paired with spatially resolved cell numbers obtained by PicoGreen double-stranded DNA quantification. Six regression algorithms were compared under identical conditions; a multilayer perceptron (MLP) gave the lowest mean absolute error (13.2%). Permutation importance ranked flow velocity, wall shear stress, von Mises stress and pressure as the dominant correlates of early adhesion. Parameter sweeps of perfusion rate and scaffold stiffness were model predictions rather than measured optima, and identified 1.4-1.6 mL/min and a moderately compliant scaffold as candidate operating points for experimental testing. The framework provides a transferable route to rational bioreactor and scaffold tuning with reduced empirical trial and error.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702301/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-15DOI: 10.3390/biomimetics11100741
Abdullah Alabdulatif
{"title":"Biological Complexity and Cryptographic Secrecy: Reproducible Key Recovery and a Structured Evidence Audit of Bio-Derived Key Schemes.","authors":"Abdullah Alabdulatif","doi":"10.3390/biomimetics11100741","DOIUrl":"10.3390/biomimetics11100741","url":null,"abstract":"<p><p>Bio-derived key schemes build cryptographic keys from biological material: protein structures, DNA and RNA sequences, amino-acid chains, proteinoid assemblies. This paper audits the class against a protocol drafted prospectively, refined during a pre-specified pilot, and frozen before the main screening and extraction, and tests its central claim empirically. No audited study reported a defensible attacker-conditioned min-entropy for its complete secret; the key-space figures instead measured an output or encoding-table size, a descriptor length, or a parameter-range product. A structured evidence audit of twelve studies finds that the biological material rarely holds the secret, that the evidence concerns the randomness of outputs rather than the secrecy of keys, and that no study provided a publicly linked runnable implementation. On three schemes (the two whose secret is a selection from a public set and one encoding scheme) we recover keys from the first two and bound the entropy of the third. The executed recoveries covered a 647-protein candidate set (work factor about 9.3 search-space bits) and a single-reference DNA position search (about 12 bits); a broader eight-reference analysis gives a projected search space upper bound of at most about 39 bits once a 16-bit auxiliary key is included, and was not exhaustively executed. Reported key spaces range up to millions of bits, and for the encoding scheme the input entropy is left unestablished. These results apply to schemes whose secret is public, a selection from public data, or an encoding; they do not extend to a scheme built on a genuinely private biological sample, a noisy physical measurement, or a properly constructed fuzzy extractor. We contribute a taxonomy of biological sources by role, three security observations, the audit and the recovery and accounting experiments, and a design checklist. Biological complexity does not by itself imply cryptographic secrecy.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702405/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261746","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-15DOI: 10.3390/biomimetics11100744
Hüseyin Yanık
{"title":"Homeostasis-Inspired, Quality-Adaptive Preprocessing for Multi-Label ECG Diagnosis: A Pre-Registered Multi-Objective Evolutionary Study with Negative Findings and Falsifiable Design Principles.","authors":"Hüseyin Yanık","doi":"10.3390/biomimetics11100744","DOIUrl":"10.3390/biomimetics11100744","url":null,"abstract":"<p><p>Deep-learning electrocardiogram (ECG) classifiers usually apply one fixed preprocessing filter to every recording, although signal quality varies. Inspired by physiological homeostasis, ECG preprocessing was formulated as a sensor-controller-effector loop: a weakly supervised quality sensor assigns each 12-lead record to a quality tier, and a tier-gated policy, shaped by Non-dominated Sorting Genetic Algorithm II (NSGA-II) on the validation fold, selects a filtering operator and a calibration temperature for a frozen compact Convolutional Neural Networks (CNN). In a pre-registered evaluation on PTB-XL v1.0.3 (21,388 records), the adaptive policy did not outperform the best fixed band-pass filter in discrimination (test macro area under the receiver operating characteristic curve (AUROC 0.9225 vs. 0.9231), worst-tier performance, or calibration, on native data or on any of 15 synthetic and real (MIT-BIH Noise Stress Test) noise benches. Like the fixed filter, it clearly outperformed unfiltered processing under baseline wander (ΔAUROC +0.0047, Holm-adjusted <i>p</i> < 0.001). Post hoc analyses explain these negative results: the operator bank offered no tier-dependent trade-off, the worst-tier objective was rarely binding, the sensor saturated because its tier boundaries were fitted on clean data, the sensor cost exceeded the filtering it saved, and the selected operators were unstable across resamples, while the qualitative rule \"filter only the disturbed tier\" was stable. Of four predictions derived from homeostasis, one was supported, two were supported only partially, and one was refuted. This study distills six falsifiable design principles for bio-inspired adaptive signal processing.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702186/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-14DOI: 10.3390/biomimetics11100737
Ezgi Yılmaz, Mehmet Sair Akkam
{"title":"Certification Criteria and Bio-Inspired Functional Themes: LEED v5 and BREEAM V7 Profiles with a Bounded YeS-TR Audit.","authors":"Ezgi Yılmaz, Mehmet Sair Akkam","doi":"10.3390/biomimetics11100737","DOIUrl":"10.3390/biomimetics11100737","url":null,"abstract":"<p><p>This study examines whether green building rating system architectures formally recognize functions associated with bio-inspired design. It compares the complete LEED v5 BD+C New Construction dataset (49 criteria; 110 points) and office-applicable BREEAM International New Construction V7 (51 issues; 165 standard credits). A Biomimetic Readiness Lens (BRL) provisionally regroups the eleven-domain Harmonized Sustainability Taxonomy into six interpretive dimensions. The mapping is deterministic and exploratory; it adds an interpretive synthesis beyond the underlying sustainability-domain results but is not independently validated and does not measure biological emulation, building performance, or design outcomes. Coverage, scoring incentives, minimum requirements, and qualitative localization are analyzed separately. Results show that resource-related criteria are prominent in both complete systems, while their credit allocation and minimum-standard structures differ. Independent recoding of all 100 items found complete agreement on primary HST domains and limited agreement on optional secondary codes; all 54 secondary differences were subsequently adjudicated under a source-based coding rule. The reconciled classifications retain 34 optional secondary assignments. The initial reliability estimates are reported separately from post-recoding adjudication. Document comparability and national source coverage also constrain interpretation. The BRL should be read as a provisional analytical crosswalk, not as a validated measure of biomimetic readiness.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702249/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261841","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-14DOI: 10.3390/biomimetics11100739
Lijuan Ren, Hang Chen, Jun Yang, Lei He, Dan Tang
{"title":"HT-Net: A Task-Specific Dual-Branch Network for Tumor Segmentation in Histopathological Images.","authors":"Lijuan Ren, Hang Chen, Jun Yang, Lei He, Dan Tang","doi":"10.3390/biomimetics11100739","DOIUrl":"10.3390/biomimetics11100739","url":null,"abstract":"<p><p>Breast cancer, the most prevalent malignancy among women globally, requires precise assessment of lymph node metastases in whole-slide images (WSIs) for accurate clinical staging and therapeutic planning. However, manual assessment is laborious and prone to inter-observer variability. Inspired by biomimetic principles of functional specialization and complementary visual perception, we propose Head-Tail Network (HT-Net), a novel dual-branch encoder network for histopathological image segmentation. HT-Net employs a HeadBranch with Squeeze-and-Excitation (SE) attention to prioritize background feature selection and a TailBranch with a novel Hybrid Multi-Scale Attention (HMSA) module to resolve complex tumor boundaries. Features from both branches are integrated through a spatially adaptive fusion module and reconstructed using a U-Net-style decoder. A task-specific supervision loss further enforces functional specialization between branches and mitigates class imbalance. Experiments demonstrate that HT-Net outperforms U-Net, UNet++, UCTransNet, VM-UNet, and nnWNet. On Camelyon16, HT-Net achieved 89.37% mDice and 82.05% mIoU, surpassing nnWNet by 0.86 and 1.08 percentage points, respectively. On Glas, it achieved 93.04% mDice and 86.95% mIoU, exceeding nnWNet by 0.30 and 0.47 percentage points. These results demonstrate the effectiveness of the biomimetic-inspired functional specialization strategy for automated histopathological image segmentation.</p>","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702355/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
BiomimeticsPub Date : 2026-10-14DOI: 10.3390/biomimetics11100738
Cebrail Barut, İnanç Özgen
{"title":"Bio-Inspired Swarm Optimization for Interpretable Classification of <i>Tropinota hirta</i> Abundance Using Partial SMOTE and XGBoost.","authors":"Cebrail Barut, İnanç Özgen","doi":"10.3390/biomimetics11100738","DOIUrl":"10.3390/biomimetics11100738","url":null,"abstract":"<p><p>Classifying blue basin trap captures of <i>Tropinota hirta</i> according to environmental conditions can support timely and data-driven management of this flower-feeding pest in almond orchards; however, imbalanced class distributions pose important challenges for machine-learning models. This study proposes an interpretable PSO-XGBoost with Partial SMOTE framework for classifying <i>T. hirta</i> abundance levels derived from blue basin trap captures using temporal and meteorological variables. From a biomimetic perspective, Particle Swarm Optimization (PSO) is a bio-inspired optimization algorithm that emulates the collective search and information-sharing behavior observed in natural swarms, translating principles of biological collective intelligence into a computational optimization process. The dataset consisted of 256 trap observations categorized into High (H), Low (L), and Medium (M) abundance classes, with 180, 55, and 21 observations, respectively, indicating substantial class imbalance. Eleven temporal and meteorological variables were used as predictors, while PestCount, representing the observed number of <i>T. hirta</i> adults captured in the blue traps and used to derive the target classes, was excluded from the predictor set to prevent target leakage. The proposed framework combines controlled minority-class oversampling using Partial SMOTE, bio-inspired PSO for XGBoost hyperparameter optimization, and rule extraction for model interpretability. A Partial-SMOTE ratio ablation analysis identified L = 0.80 and M = 0.40 as the most favorable class-specific oversampling configuration, avoiding complete class balancing while improving overall and class-sensitive predictive performance. Performance was evaluated using stratified five-fold cross-validation, with oversampling applied exclusively to the training folds to prevent information leakage. The proposed method achieved a mean Accuracy of 89.45% and a mean Weighted F1-score of 88.79%, representing the best overall performance among the evaluated methods. The ablation analysis showed that PSO-based optimization improved conventional XGBoost, while the proposed Partial-SMOTE configuration outperformed both class-weighted XGBoost and PSO-XGBoost without oversampling. Robustness was further assessed using 10-times repeated stratified five-fold cross-validation, where the proposed method achieved 87.51 ± 3.75% Accuracy, 86.94 ± 3.71% Weighted F1-score, and 71.60 ± 7.40% Macro F1-score. Friedman and post-hoc Wilcoxon signed-rank tests with Holm correction confirmed statistically significant performance differences in favor of the proposed method. In addition, human-readable IF-THEN rules with bounded temporal and meteorological conditions were extracted from the optimized model to improve interpretability. The results indicate that integrating a biomimetic swarm-intelligence optimization strategy with controlled oversampling, blue-trap monitoring, and interpretable machine learning","PeriodicalId":8907,"journal":{"name":"Biomimetics","volume":"11 10","pages":""},"PeriodicalIF":4.2,"publicationDate":"2026-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13702523/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149261755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}