Search Results (2,322)

Survival-rule induction is commonly guided by log-rank separation, whereas some prognostic questions target conditional future lifetime. We propose a directional finite-horizon mean residual life (MRL) criterion for survival-rule induction under right censoring, combining normalized subgroup support with a survival-weighted restricted-MRL discrepancy. The framework includes favorable and adverse objectives, training-only horizon selection and tuning, overlapping-rule prediction, and finite-candidate plug-in consistency. Across nine simulation scenarios (200 replications each), MRL recovered the true subgroup partition more accurately under delayed benefit, crossing hazards, delayed benefit with 60% censoring, and small-sample crossing; log-rank was stronger under proportional hazards, early-only effects, rare subgroups, and the adverse stress test. In repeated nested analyses, Cox proportional hazards achieved the lowest mean integrated Brier score (IBS) in WHAS100 (0.1797) and malignant melanoma (0.1300); controlled log-rank also yielded lower IBS than MRL (0.1926 vs. 0.2038 and 0.1358 vs. 0.1404, respectively). MRL nevertheless identified different conditional-lifetime structures. These results position MRL-guided rules as an estimand-specific complement to hazard-oriented methods rather than a universally superior predictor. Full article

In post-mortem genetic identification, teeth represent one of the most reliable sources of nuclear DNA when other biological samples are unavailable, owing to their ability to protect genetic material from autolysis, microbial degradation, and environmental insults. The present study describes a forensic veterinary case report involving a Staffordshire Bull Terrier litter in which the appearance of blue-coated offspring from a phenotypically black sire and a blue dam prompted parentage verification. The sire died before formal ante-mortem sampling could be carried out. In accordance with Regulation (EC) No. 1069/2009, Article 15, the carcass was interred in the owner’s garden. Before burial, two teeth were collected as the sole biological material available for identification by the official veterinarian, ensuring an unambiguous chain of custody from death certification to laboratory analysis. DNA extraction was performed using two independent protocols: Method A, a silica-membrane-based purification (QIAamp DNA Kit, Qiagen), applied to dental pulp, and Method B, a decalcification-based workflow (T-Bone Ex Kit) followed by DNA extraction using EZ2 Connect Instruments (Qiagen). Quantification of the DNA samples obtained were done by fluorometry. Short Tandem Repeat (STR) genotyping was performed using the Canine Genotypes™ Panel 1.1 by Applied Biosystems (Thermo Fisher Scientific, Waltham, MA, USA) targeting the 18 autosomal ISAG-recommended microsatellite loci. Parentage assignments were carried out against reference STR profiles obtained from peripheral blood of the dam and offspring. The two methods, considering a different biological matrix and different operating conditions, allowed us to obtain the STR profile, but Method B showed quality electropherograms, improving the confidence and interpretation of all the markers analyzed. The differences observed between the two approaches may reflect a combination of factors, including the different dental matrices analysed, the quantity and quality of the starting material, tooth identity and tissue preservation, and differences inherent to the extraction procedures themselves; therefore, in this case, the more complete and clearer STR profile obtained with Method B should be considered a case-specific observation rather than evidence of a direct effect of the extraction method. Overall, regardless of the DNA extraction method used, the results did not exclude the biological paternity of the blue-mantled offspring. Furthermore, this case report provides practical methodological guidance for DNA recovery from canine teeth and supports their use as a reliable source of post-mortem DNA for STR-based parentage testing. Full article

Background: Cucurbita ficifolia (figleaf gourd) is an agriculturally useful cucurbit germplasm and rootstock, but its mitochondrial genome has not been examined in a comparative family-level framework. We aimed to characterize its mitochondrial reference sequence and evaluate coding composition, repetitive DNA, plastid homology, bioinformatic RNA-editing predictions, phylogenetic placement, and genome-scale synteny. Methods: PacBio Revio sequencing generated 933,158 high-fidelity (HiFi) reads totaling 14.294 Gb, with a read N50 of 15.6 kb. The 784,544-bp mitochondrial reference (GenBank accession PZ823090) was assembled with Oatk and evaluated using graph topology, competitive read-back, and junction-spanning alignments. Codon usage, repeats, mitochondrial–plastid homologous regions, bioinformatic C-to-U RNA-editing predictions, conserved-gene phylogeny, and whole-mitogenome synteny were analyzed. Results: The reference had 43.09% GC content and contained 38 distinct protein-coding genes represented by 42 loci, 26 tRNA types represented by 37 loci, and three rRNAs. Competitive read-back retained 40,971 primary mitochondrial alignments (MAPQ ≥ 20), yielding a mean depth of 547.26× and 100% breadth at ≥100×. The terminal-to-start adjacency was supported by 366 HiFi reads with ≥2-kb anchors on each side. We identified 322 simple-sequence repeats, 115 tandem repeats, and 2374 dispersed repeat pairs. For the largest 392-bp direct repeat, the two native junctions were supported by 268 and 308 reads, whereas the two reciprocal junctions were supported by one and zero reads. Ninety-one significant plastid matches collapsed to 73 mitochondrial loci and covered 74,784 unique mitochondrial bases (9.53%). Deepred-Mt generated 494 bioinformatic C-to-U RNA-editing predictions in 37 genes, and phylogenomics placed C. ficifolia as sister to the C. pepo–C. maxima pair with maximal support despite extensive synteny turnover. Conclusions: The C. ficifolia mitogenome combines a conserved coding repertoire with abundant short repeats, appreciable plastid homology, and extensive rearrangement. Read evidence supports the submitted terminal-to-start adjacency but does not establish a unique or predominant circular molecule in vivo; likewise, the largest repeat showed no robust reciprocal-junction signal. All RNA-editing sites reported here are bioinformatic predictions rather than transcript–validated events. Full article

This study addresses the lack of quantitative understanding regarding vent-gas combustion feedback in thermal runaway propagation of manganese-based prismatic battery modules. We develop a coupled multiphysics model integrating solid heat transfer, runaway kinetics, gas ejection, and turbulent combustion, informed by constant-volume reactor tests on a single cell that yield a total gas release of 10.6 mol per cell, with H₂ and CO comprising about 49% of combustibles. Model predictions are validated against three-cell module propagation experiments, capturing sequential failure with simulated inter-cell intervals matching experimental repeats that range from 71 to 105 s. The combustion model assumes auto-ignition-based initiation and therefore does not address the stochastic ignition delays observed in the experiments. It is intended for quantifying propagation acceleration once combustion has commenced. Quantitative heat-flow analysis reveals that gas-phase convective and radiative heating contributes substantially to the total heat flux on adjacent cells during venting, and direct comparison between simulations with and without combustion shows that combustion reduces inter-cell propagation intervals by 8–11%, confirming that combustion actively accelerates propagation. Simulations further resolve the combustion zone extending up to 0.4 m laterally. This framework provides a predictive tool and mechanistic basis for vent-gas management and flame-mitigation strategies in battery energy storage systems. Full article

Persistent open-vocabulary navigation enables robots to search for sequential language-specified objects using reusable visual semantic evidence. On edge GPUs, the pipeline is constrained by foundation-model inference, semantic projection, persistent map movement, frontier processing, and repeated target detection. Using OneMap as a controlled reference, we reorganize this workload through compiled mixed-precision perception, direct-to-map confidence-weighted feature aggregation, accelerator-resident semantic maps, compiled navigation operations, and language-conditioned detector scheduling. The resulting dataflow removes the dense image resolution feature intermediate and schedules target confirmation detection from current-view target similarity. On the complete Habitat-Matterport 3D benchmarks, our implementation retains 92.6% and 96.1% of the reproduced OneMap single- and multi-object success rates. Across 36 paired hardware-in-the-loop comparisons on a Jetson AGX Orin at four power modes, it achieves geometric mean speedups of 4.64× for on-device computation and 3.55× for the complete episode duration. It also reduces peak system random access memory (RAM) by 51.21% and a summed board-rail energy proxy by 75.32% on average while preserving every paired outcome. These results establish the coordinated dataflow and current-view target-conditioned execution as effective mechanisms for efficient persistent foundation-model navigation on power-constrained edge robots. Full article

Background/Objectives: Postprandial glycemic responses vary markedly among individuals, but it remains unclear whether the glycemic effect of incremental carbohydrate exposure is systematically modified by insulin resistance. We examined whether insulin resistance alters the carbohydrate–postprandial glucose dose-response across distinct meal contexts. Methods: We performed a secondary repeated-measures analysis of the publicly available CGMacros cohort. Forty-five adults spanning normoglycemia, prediabetes, and type 2 diabetes wore blinded FreeStyle Libre Pro and Dexcom G6 Pro continuous glucose monitors while recording meals for approximately 10 days. The primary outcome was 2-h incremental area under the glucose curve (iAUC). Mixed-effects models tested carbohydrate × HOMA-IR interactions, adjusted for protein, fat, fiber, premeal glucose, age, sex, and BMI. Breakfast was the primary structured context, lunch provided within-cohort contextual confirmation, dinner was a free-living contrast, and Dexcom analyses assessed cross-device robustness. Results: The primary Libre analysis included 423 breakfasts and 414 lunches from 44 participants. Higher HOMA-IR significantly amplified the iAUC associated with each 10-g increment in carbohydrate during breakfast (interaction β = 178.5 mg·min/dL per 1-SD higher HOMA-IR; 95% CI 79.5–277.5; p < 0.001) and lunch (β = 109.8; 95% CI 34.2–185.3; p = 0.004). The direction was supported with Dexcom during breakfast (β = 165.3; 95% CI 45.0–285.6; p = 0.007) and lunch (β = 87.9; 95% CI 2.6–173.1; p = 0.043) and remained robust in GEE, participant fixed-effects, random-slope, activity-adjusted, and other sensitivity analyses. No positive interaction was observed during dinner. Conclusions: In this observational secondary analysis, insulin resistance modified the glycemic impact of carbohydrate across two structured meal contexts, with supportive cross-device evidence. These findings are hypothesis-generating and do not establish HOMA-IR-based dietary prescriptions; prospective external validation is required. Full article

Background: Helicobacter pylori infection is a common chronic bacterial infection and an established cause of gastritis, peptic ulcer disease, and gastric cancer. Environmental and host metal status may interact with gastric inflammatory and microbial processes, and infection-related changes in gastric acidity, mucosal permeability, and metal handling may in turn alter local tissue concentrations. This study compared systemic and gastric tissue concentrations of seven elements in adults with histopathologically confirmed H. pylori infection and uninfected controls. Methods: This single-center case–control study included 110 adults undergoing upper gastrointestinal endoscopy (55 H. pylori-positive and 55 H. pylori-negative). H. pylori status was established by hematoxylin–eosin and modified Giemsa staining. Whole blood collected in purple-top K₂EDTA tubes was analyzed for As, Pb, Cd, and Sn; serum obtained from yellow-top serum-separator tubes containing clot activator and separator gel was analyzed for Al, Co, and Ni. A separate fresh antral biopsy, not the formalin-fixed histology specimen, was used for ICP-MS; tissue concentrations were normalized to recorded dry weight. Between-group metal comparisons were corrected as one family of 14 tests using Benjamini–Hochberg FDR. Prespecified age- and sex-adjusted models and an exploratory multivariable model were supplemented by ROC, PCA, and internally cross-validated PLS-DA analyses. Results: The H. pylori-positive group was older than the negative group (55.65 ± 15.01 vs. 52.42 ± 12.53 years). Blood aluminum was nominally higher in the positive group (p = 0.032; q = 0.112). In gastric tissue, aluminum was higher [0.23 (0.13–13.40) vs. 0.13 (0.10–0.23) μg/g dry weight; p < 0.001; q = 0.002], whereas arsenic was lower [1.35 (0.54–2.02) vs. 1.92 (1.47–2.43) μg/g dry weight; p < 0.001; q = 0.003] in H. pylori-positive participants. Tissue cobalt and blood/tissue nickel differences did not remain significant after FDR correction. In prespecified age- and sex-adjusted models, tissue arsenic showed an inverse association (OR = 0.431, 95% CI 0.263–0.705; p = 0.0008), whereas tissue aluminum showed a positive association (OR = 4.957, 95% CI 1.629–15.085; p = 0.0048). In the exploratory joint model, only tissue aluminum remained significant (adjusted OR = 4.016, 95% CI 1.119–14.417; p = 0.033). The joint model had an apparent within-cohort AUC of 0.774; repeated five-fold cross-validated PLS-DA yielded an AUC of 0.707. Conclusions: Histopathologically confirmed H. pylori status was associated with compartment-specific metal concentration patterns, particularly higher tissue aluminum and lower tissue arsenic. Because the study was cross-sectional and measured total concentrations, these findings do not establish environmental exposure, temporal accumulation, toxicity, or a causal direction. The regression and multivariate findings are exploratory and require validation in independent cohorts with standardized tissue sampling, arsenic speciation, exposure assessment, and quantitative gastric pathology. Full article

Jujube is an important fruit tree and one of the five major economic forest tree species native to China, possessing high nutritional and medicinal value. It plays a key supporting role in the efficient utilization of marginal land resources—such as mountainous, sandy, saline-alkali, and drought-prone areas—as well as in the revitalization of rural industries. Molecular marker technology, with its advantages of stability, accuracy, and efficiency, has become an important tool in jujube genetic breeding research and is widely applied. This article summarizes molecular markers based on three generations of technological intergenerational systems: the first generation of hybridization-based markers (Restriction Fragment Length Polymorphism, RFLP), the second generation of PCR-based markers (represented by Simple Sequence Repeat, SSR), and the third generation of high-throughput sequencing markers (Single Nucleotide Polymorphism, SNP, Genotyping-by-Sequencing, GBS, etc.). A comprehensive classification system based on technical principles, polymorphism sources, and other dimensions is constructed to clearly explain the driving forces and development trends of molecular marker system evolution in jujube tree research, and to clarify the selection criteria and adaptation strategies of molecular markers. Research has found that the second-generation molecular marker SSR is the fundamental core tool for standardized identification of jujube germplasm resources and genetic analysis of low-budget populations. The third-generation molecular markers represented by SNPs and InDel are high-density maps, Genome-Wide Association Study(GWAS), Genomic selection, and other modern precision breeding core carriers, forming a layered complementary technology system, and are gradually becoming the mainstream of research. This paper summarizes the application progress of molecular markers in the precise identification of jujube germplasm resources, analysis of genetic diversity, determination of genetic relationships, construction of high-density genetic maps, genome-wide association analysis, functional gene mapping, and tracing of domestication and evolution. This article integrates all existing research using the unified scientific framework of “technology defect driven tagging iteration”, analyzes the internal logic of the evolution and replacement of different tagging systems, the inherent limitations of early tagging, the existing problems in current research, and discusses future research directions, aiming to provide a reference for the scientific and efficient application of molecular markers in jujube and to promote the improvement and upgrading of the jujube molecular marker-assisted breeding technology system. Full article

This in vitro study evaluated the effects of layer thickness, photoactivation protocol, glycerin application, and polishing on top-surface roughness and microhardness. Sixty specimens were prepared from bulk-fill (Tetric PowerFill) and conventional (IPS Empress Direct) composites. Bulk-fill specimens were 2 or 4 mm thick and photoactivated with a conventional (1200 mW/cm², 20 s) or high-speed (3000 mW/cm², 3 s) protocol, giving nominal radiant exposures of 24 and 9 J/cm²; conventional specimens were 2 mm thick and photoactivated using the conventional protocol. Glycerin was applied or omitted; all specimens were polymerized beneath a polyester strip, with no air-exposed group. Roughness and Vickers microhardness were measured exclusively on the top surface, before and after polishing, and analyzed using General Linear Models for repeated measures (α = 0.05). In the bulk-fill composite, polishing reduced roughness and increased microhardness, with significant polishing × glycerin interactions for both outcomes. In the conventional composite, glycerin increased microhardness and roughness before polishing. Polishing increased microhardness, and a significant polishing × glycerin interaction was observed for roughness. The 20 s protocol produced significantly higher top-surface microhardness than the 3 s protocol, although the difference was small (2.58 VHN); because irradiance and exposure time differed simultaneously, their individual contributions cannot be distinguished. Layer thickness did not significantly affect the top-surface properties evaluated; this cannot be extrapolated to deeper regions or taken as evidence of equivalence between the 2 and 4 mm increments. In the condition in which both composites were treated identically, the effects of glycerin and of polishing differed significantly between the materials for both outcomes; for roughness, the difference in glycerin effects further depended on polishing. With no air-exposed controls and no artificial aging, the findings are restricted to the immediate top-surface response under the conditions evaluated. Full article

Operational redundancies in Nigeria’s airport logistics, including duplicated clearance procedures, fragmented data systems, and repeated physical inspections, generate substantial inefficiencies that hamper service delivery, increase costs, and raise sustainability concerns. Despite growing interest in blockchain technology (BCT) as a logistics innovation, empirical evidence from Sub-Saharan African aviation contexts is largely limited. This study examines how stakeholders perceive BCT as a potential mechanism for reducing procedural, informational, and structural redundancies across Nigerian airport operations. A qualitative research design was employed, drawing on 45 semi-structured interviews with airport managers, airline officials, regulatory bodies, and aviation experts. Data were analyzed using Leximancer, yielding eight thematic clusters across 115 concepts. To augment thematic rigor, three quantitative measures—Normalized Pointwise Mutual Information (NPMI), Jaccard similarity, and degree centrality—were applied to the exported Concept Co-occurrence Matrix (CCMatrix; 6555 pairwise relationships). The blockchain theme exhibited the highest intra-theme cohesion (mean NPMI = 0.615) and the second highest normalized centrality (0.638), suggesting that it functions as a semantically tight, cross-cutting concept in participants’ discourse. The strongest concept pair, inefficiency–reliability (NPMI = 0.989; Jaccard = 0.909), is consistent with participants’ recurring association of service failures with procedurally generated redundancies. Stakeholders perceived BCT as potentially addressing these redundancies through shared ledger architectures, smart contract automation, and consolidated identity management, with anticipated sustainability co-benefits across energy, documentation and processing domains. The study offers an exploratory, contextualized application of the Technology–Organization–Environment (TOE) framework, the Unified Theory of Acceptance and Use of Technology (UTAUT), and the Dynamic Capabilities Framework (DCF) to a resource-constrained developing-country aviation context and proposes a stakeholder-informed, phased BCT adoption framework—not yet tested through implementation—with direct policy and managerial implications for the Nigerian airport administration. Full article

Financial time-series forecasting remains challenging because of high volatility, nonlinear market dynamics, and the growing volume of heterogeneous information available to market participants. This study investigates Bitcoin forecasting in a big-data setting by combining high-frequency market data aggregated to hourly forecasting features, news headlines, Bitcoin on-chain variables, and broader macro-financial indicators. Rather than treating headline sentiment as a predefined categorical property, we construct continuous headline-conditioned market-impact labels from short-horizon Bitcoin price responses, directional volume imbalance, and volatility conditions. A chronology-controlled expanding-window FinBERT procedure generates scores without reusing each headline’s own future-derived target. The scores are integrated into multivariate Bitcoin forecasting using iTransformer, with LSTM as a benchmark. Evaluation uses repeated runs, benchmarks, Diebold–Mariano tests, backtesting, and forecast-free momentum controls. The headline-derived signal exhibits a measurable but temporally heterogeneous association with subsequent Bitcoin movements. Adding the headline score yields small, statistically non-significant error reductions for iTransformer, whereas it significantly worsens LSTM forecasts. Backtesting shows no consistent improvement in terminal portfolio value, but the headline feature alters the risk–return profile in several strategy configurations. Overall, the study provides a chronology-aware evaluation of whether headline-conditioned market-impact signals add predictive or economic value, while acknowledging residual dependence from exploratory iTransformer architecture selection. Full article

Efficient evaluation of aptamer candidates after systematic evolution of ligands by exponential enrichment (SELEX) remains limited by the time and cost required for direct kinetic characterization. We developed a centrifugal microspin filtration (MSF)-based MSF–k_off assay for retention-based, dissociation-sensitive pre-screening of aptamer candidates. Repeated washing preferentially removes rapidly dissociating aptamers from target-immobilized magnetic beads, and the retained aptamers are quantified by quantitative polymerase chain reaction and expressed as Wash 0-normalized retention (W0-NRT). In a thrombin aptamer panel, W0-NRT showed a strong inverse relationship with SPR-derived k_off. A preliminary analysis of streptavidin aptamers likewise showed a W0-NRT-based ranking consistent with the literature-reported k_off values, although the two target systems exhibited distinct empirical relationships. In a wash-based thrombin sandwich assay, two affinity-matched sensor aptamers differing 1.76-fold in k_off were ranked in the same order by post-wash retention as by W0-NRT. The agreement was in rank order rather than in the magnitude of retention. The parallel microspin filter format was also estimated to offer advantages in analysis time and assay-related cost. The MSF–k_off assay therefore provides a simple post-SELEX approach for prioritizing candidates with higher retention under dissociation-sensitive washing before detailed kinetic characterization. Full article

Ultrasonic pulse-echo imaging is one of the most widely used non-destructive evaluation (NDE) modalities for monitoring the structural integrity of reinforced-concrete civil infrastructure such as bridge decks, tunnel linings, and dam walls. In this acquisition geometry, a single low-frequency transducer is mechanically raster-scanned over the accessible top surface of the specimen and records one A-scan per scan position, simultaneously serving as transmitter and receiver. However, commonly used reconstruction algorithms such as the Synthetic Aperture Focusing Technique (SAFT) and Reverse Time Migration (RTM) tend to produce reconstructions of limited quality on this class of data because they do not adequately model the round-trip propagation kernel that is specific to the monostatic geometry, they do not separate the strong near-surface direct-arrival reflection from the bulk image, and they do not account for the persistent aggregate-induced clutter that contaminates every A-scan in concrete media. In this paper, we propose a clutter-aware reconstruction method for monostatic ultrasound acquisition (CARMA), whose main innovation is the joint integration of a monostatic-specific round-trip propagation model, a dedicated near-surface direct-arrival subspace, and a data-adaptive low-rank clutter subspace within a unified model-based reconstruction framework. Unlike existing reconstruction approaches, CARMA explicitly accounts for the co-located transmit–receive geometry through a squared-cosine round-trip directivity model while simultaneously separating scan-dependent direct-arrival contributions and aggregate-induced clutter from the desired reflectivity image. To verify the method under fully controlled and repeatable conditions, we generate intensive, physically realistic full-wave simulations with the k-Wave pseudo-spectral acoustic solver that reproduce a representative civil-infrastructure inspection scenario: three reinforced-concrete specimens with a stepped back wall of varying thickness, ten embedded ground-truth defects spanning steel tendon ducts and low-impedance polystyrene inclusions, a monostatic raster-scanned pulse-echo acquisition, and randomly distributed aggregate scatterers that reproduce the clutter of real concrete. Results on these intensive k-Wave simulations indicate that CARMA reconstruction yields approximately $2\times$ lower localization error than RTM and approximately $4\times$ lower localization error than SAFT, while recovering the deepest embedded defect with substantially better localization and contrast than the comparison methods. Full article

Background and Objectives: Low-volume plyometric conditioning may elicit post-activation performance enhancement (PAPE), but the early response to pogo-jump exercise in adolescent basketball players remains uncertain. This randomized crossover pilot study estimated the effect of adding 24 bilateral pogo jumps to a standardized warm-up on countermovement-jump (CMJ) and drop-jump (DJ) height, compared with time-matched low-intensity walking. Methods: Twenty-seven male basketball players were recruited, and 21 (age, 16.55 ± 0.88 years) completed both conditions. In randomized order, separated by a 7-day washout, participants performed 3 sets of 8 bilateral pogo jumps with 1 min between sets or approximately 3 min of low-intensity walking. Post-condition jump assessments began after a 1 min recovery and comprised three consecutive CMJs, a 30 s transition, and three consecutive DJs, with all measurements completed within approximately 2 min after each condition. The highest valid trial was retained for each jump type. Period-adjusted treatment effects were estimated using separate linear mixed-effects models, with repeated-measures ANOVA and Wilcoxon analyses retained as sensitivity analyses. Results: The period-adjusted between-condition difference in change was −0.89 cm for CMJ (95% CI, −2.92 to 1.14; p = 0.389) and 0.34 cm for DJ (95% CI, −1.77 to 2.45; p = 0.753). Sensitivity analyses yielded the same overall interpretation. All completers performed the protocol, and perceived exertion was 11.4 ± 1.2. The pogo-jump protocol showed favorable immediate tolerability, with all completers reporting 0/10 anterior-knee pain immediately after the condition. Conclusions: The pilot estimates did not indicate a detectable group-level improvement in CMJ or DJ height during the early post-conditioning assessment. However, the confidence intervals were compatible with potentially meaningful effects in either direction. The small sample, single dose, single early assessment window, and variable DJ reliability limit the precision and generalizability of the findings. Full article

Under high renewable-energy penetration, regional power systems face increasing operational variability caused by wind and photovoltaic generation. Capacity planning for multi-energy complementary systems must therefore consider both investment economy and operation-related reliability costs under practical grid-access constraints. This study proposes a substation-constrained bi-level capacity-planning framework for a regional multi-energy power system in Ningxia. In the upper level, the installed capacities of wind, photovoltaic, thermal, and energy-storage resources are optimized by minimizing the annualized comprehensive cost subject to regional substation capacity limits. In the lower level, a mixed-integer linear programming dispatch model evaluates the operating cost of each candidate capacity scheme under representative seasonal scenarios, including thermal fuel cost, net grid-exchange cost, and reliability-related penalty costs associated with load curtailment, reserve shortage, and power imbalance. An Improved Whale Optimization Algorithm is used as the outer planning optimizer and is coupled with the lower-level MILP dispatch model. The case study uses normalized hourly load, wind power, and photovoltaic power data from Ningxia to construct an equivalent three-region planning system. The results show that the proposed bi-level framework reduces the annualized comprehensive cost by 28.3% compared with the single-level model, with the minimum total system cost reaching 2.71 × 10⁸ CNY. Compared with the standard WOA, PSO, and GA under the same case setting, the proposed IWOA obtains a lower objective value and smaller variation across repeated runs. The reliability-related terms in the model are interpreted as economic operational proxies rather than direct probabilistic reliability indices. The proposed framework provides a planning-oriented method for evaluating multi-energy capacity allocation under substation capacity constraints and representative operational scenarios. Full article