Lumbar assist suits are commonly evaluated based on reductions in muscle activity; however, such reductions do not necessarily correspond to decreases in intervertebral loading. This study developed a two-dimensional sagittal-plane static rigid-link model to evaluate relative changes in lumbar intervertebral loads during assist suit use. The model incorporated subject-specific spinal alignment based on age, sex, and height and estimated compressive force, shear force, and external flexion moment at each intervertebral level from L1/L2 to L5/S1. Three conceptual assist conditions—shoulder-traction assistance, extensor-muscle assistance, and lumbar stabilization—were compared at L4/L5 and L5/S1. The assist parameters were calibrated to reduce the L5/S1 ES-group force by 20% at 60° trunk flexion. Under shoulder-traction assistance, compressive force, shear force, and external flexion moment decreased at both L4/L5 and L5/S1. Under extensor-muscle assistance, erector spinae force decreased as prescribed, whereas the three intervertebral load components remained unchanged. Lumbar stabilization reduced compressive force and external flexion moment but slightly increased shear force. These findings demonstrate that similar reductions in erector spinae force can produce different changes in intervertebral loads depending on the mechanical assistance principle, highlighting the importance of considering multiple load components when evaluating and designing lumbar assist systems.
Federated learning (FL) supports collaborative edge intelligence without centralizing raw data, but non-IID data, Byzantine attacks, and limited client energy complicate IoT deployment. We study two complementary server-side mechanisms: TEA-FL, which couples consistency-based trust with residual energy in client selection and aggregation, and ARL-FL, which separates risk/lifetime-aware scheduling from pressure-adaptive robust aggregation. A matched two-domain protocol distinguishes full-system PAMAP2 evidence from aggregation-focused CICIoT2023 evidence and combines fold/seed-level inference with ablations, targeted sensitivities, and complete-round overhead measurements. Under the primary simulated resource model, TEA-FL achieved the highest overall PAMAP2 Macro-F1 and lowest normalized energy, but reduced participation fairness and did not complete every nominal-horizon run. ARL-FL preserved the largest active-client population and completed all PAMAP2 runs, but failed under the strongest sign-flip attack. Focused controls showed that trust drives TEA-FL’s sign-flip robustness, energy-aware selection drives ARL-FL’s sustainability benefit, and simpler fixed clipping-plus-trimming can be competitive under strong attack. The results support context-dependent robustness–sustainability trade-offs rather than universal superiority.
Digital cultural heritage (DCH) platforms deliver complex multimedia narratives, but their cognitive demands remain insufficiently evaluated. We recorded electroencephalography (EEG) and eye tracking (ET) from 19 valid participants (of 30 recruited) interacting with seven authentic material formats from a national public service platform. Labels were three tiers derived from material-level NASA-TLX means; modeling used leave-one-subject-out validation (LOSO) with data-driven steps fitted within training folds. The primary three-tier model, using ET with random forest, reached 0.669 accuracy (permutation p = 0.0033; 95% CI [0.595, 0.743]; macro-F1 0.66). A complementary two-tier analysis retaining all trials reached 0.737 accuracy (macro-F1 0.73). Nested LOSO with inner-fold model selection yielded three-tier accuracy 0.632 (95% CI [0.564, 0.707]; macro-F1 0.630) and two-tier accuracy 0.684 (95% CI [0.632, 0.744]; macro-F1 0.676). EEG differentiated material formats at the group level (frontal theta, p = 0.0023) but did not improve trial-level classification. Fold-internal TreeSHAP and permutation importance converged on attentional narrowing under high load: shorter saccades, longer dwell, and fewer unique gaze regions. Objective information-density measures attributed the highest load to cross-channel integration demand, not information amount. These findings support design principles for cognitively sustainable digital cultural platforms and delineate the evidentiary boundaries of small-sample multimodal load modeling.
The mechanical response of a polymeric liner in a Type IV hydrogen pressure vessel is governed by internal gas pressure, liner stiffness, nonlinear deformation, and confinement by the composite overwrap. A reduced-order analytical model is developed to describe the static and nonlinear dynamic behavior of the liner using a single generalized radial displacement. The formulation assumes cylindrical liner geometry, isothermal gas behavior, approximately constant gas mass over the mechanical response time scale, and negligible damping. The model incorporates deformation-dependent gas pressure, nonlinear restoring characteristics, effective fluid inertia, and finite clearance between the liner and overwrap. Analytical conditions for equilibrium, stability, and saddle-node bifurcation are derived. For the linear restoring model, a first integral characterizes finite-amplitude oscillations and the conditions for contact with the overwrap. Fluid inertia affects finite-amplitude motion but not static equilibrium or infinitesimal oscillations. The analysis further shows that finite-amplitude oscillations induced by an initial velocity can produce both contraction and expansion of the liner, with the diameter becoming smaller or larger than its initial value D0. Application to a representative PA11 liner demonstrates the strong effect of overwrap confinement, which substantially reduces the radial deformation. The results distinguish static and dynamic contact conditions and provide a compact analytical framework for assessing deformation, stability, and nonlinear dynamic behavior of Type IV hydrogen vessel liners.
