Journal of Progress in Engineering and Physical Science

Algorithmic Reinforcement and the Co-Evolution of User Preferences Through a Mechanistic Analysis of Diversity Loss in Recommender Systems

K. Nowak — Thu, 06 Nov 2025 00:00:00 +0000

Recommender systems have evolved into adaptive infrastructures that mediate human attention, learning, and decision-making across digital environments. This paper presents a mechanistic analysis of how algorithmic reinforcement processes co-evolve with user preferences, producing a progressive reduction in informational diversity. By conceptualizing recommendation as a coupled dynamical system, the study explains how reinforcement learning architectures internalize behavioral feedback and transform transient user interactions into long-term preference structures. The analysis identifies a recursive mechanism in which both algorithmic policies and user cognition adapt toward equilibrium states that maximize predictability and engagement at the expense of novelty. Empirical findings and theoretical models from recent reinforcement learning research are synthesized to elucidate the dynamics of diversity loss as an emergent property of co-adaptation. The paper proposes a mechanistic framework that integrates stochastic exploration, entropy regularization, and temporal reward shaping to sustain informational variety in reinforcement-driven ecosystems. This approach reconceptualizes recommender systems as co-evolutionary environments where the preservation of diversity is a structural necessity for epistemic resilience, cognitive openness, and sustainable engagement.

The Application of IoT-Empowered Intelligent Control System for High-Speed Dispersing Equipment in Textile Printing and Dyeing Additives Production

Dongmei Shi — Thu, 06 Nov 2025 00:00:00 +0000

This paper focuses on the textile printing and dyeing additives production sector, targeting the high-speed dispersing equipment, a crucial production device. A smart control system based on Internet of Things (IoT) technology has been developed and applied. By installing high-precision sensors on the equipment to real-time monitor 12 key parameters such as rotational speed and temperature, and leveraging advanced AI algorithms to dynamically optimize the stirring process, the system has been proven effective. It can reduce labor costs and reliance on skilled workers. Based on these achievements, a universal intelligent equipment upgrade framework for the textile printing and dyeing additives industry is proposed, aiming to provide references for other enterprises’ intelligent transformation within the industry. This research not only offers strong support for Wuxi Lianda Chemical’s production efficiency improvement and cost control, but also paves a new way for the sustainable development and intelligent upgrade of the whole textile printing and dyeing additives industry.

Modification of High-Precision Conductive Shielding Mylar Material and Research on Intelligent Die-Cutting Technology

Quanzhen Ding — Thu, 06 Nov 2025 00:00:00 +0000

With the rapid development of 5G communication and electronic devices in new energy vehicles, the demand for high-precision conductive shielding Mylar materials is increasing. This study successfully developed conductive shielding Mylar materials with a shielding effectiveness of 45–50 dB by modifying PET substrates using a vacuum sputtering-electroplating composite process, which is significantly higher than that of domestic counterparts (typically below 35 dB) and comparable to high-end imported materials. The adhesion test results showed that the modified material achieved a 4B level in the cross-cut test, with no detachment after 3M tape adhesion, indicating excellent adhesion. Additionally, the modified material exhibited good weather resistance, with a color difference (ΔE) of only 1.2 after 1000 hours of UVB-313 lamp irradiation.

Modular Design-Driven Lightweight Deployment Suite for SMEs’ SAP System: Development, Performance Optimization, and Industrial Validation

Qiang Fu — Thu, 06 Nov 2025 00:00:00 +0000

Against the backdrop of global digital transformation, Small and Medium-sized Enterprises (SMEs) face a paradox: they urgently need Enterprise Resource Planning (ERP) systems represented by SAP to enhance operational efficiency, yet they are constrained by limited budgets, weak technical capabilities, and low server configurations, making it difficult to adopt full-version SAP systems. This study proposes a modular design-driven lightweight SAP deployment suite to address the core pain points of high deployment costs, poor resource adaptability, and complex operations in SMEs’ SAP adoption. First, based on a demand survey of 120 manufacturing SMEs across the Yangtze River Delta, Pearl River Delta, and Bohai Rim regions, the core functions of the SAP system were decomposed into three core modules (Master Data-FI Basic Linkage, Lightweight Maintenance, Procurement-Sales Basic) and two expandable components (Inventory Early Warning, Simple Report Generation), eliminating 17 redundant sub-modules with a demand rate of less than 8%. Second, a Dynamic Resource Allocation Algorithm (DRAA) was developed, which adjusts cloud server CPU/memory resources in real time based on business volume fluctuations (e.g., order processing volume, inventory update frequency), and a Module Interface Adaptation Protocol (MIAP) was designed to achieve 99.8% data synchronization accuracy between modules with a delay of ≤5 minutes. Finally, through code compression (removing 32% of redundant ABAP code) and cache preloading technology, the module initial loading time was optimized. A 6-month controlled experiment was conducted on 30 manufacturing SMEs (covering electronics, machinery, and food industries) in three major economic belts. The results showed that compared with the traditional streamlined SAP version: (1) the annual deployment cost of the experimental group was reduced from 520,000 CNY to 198,000 CNY, with an optimization rate of 61.9% (p<0.01); (2) the deployment cycle was shortened from 45 days to 14.2 days, a reduction of 68.4% (p<0.001); (3) the peak server CPU utilization rate dropped from 80% to 42.3%, a decrease of 47.1% (p<0.01); (4) the order processing response time was reduced from 3.5 seconds to 1.1 seconds, an improvement of 68.6% (p<0.001). (Xiong, X., Zhang, X., Jiang, W., Liu, T., Liu, Y., & Liu, L., 2024) The suite has been incorporated into Accenture’s “Global SME SAP Implementation Toolkit” and promoted in 217 SMEs, creating direct economic benefits of over 65 million CNY. Future work will integrate a generative AI-driven fault diagnosis module to further reduce SMEs’ annual maintenance costs by an estimated 25-30%.

Carbon Emission Reduction Estimation and Practice of Energy-Saving Retrofit of Air-Cooling System in Thermal Power Plants

Liqin Liu — Thu, 06 Nov 2025 00:00:00 +0000

To address the critical challenge of balancing thermal power generation efficiency and low-carbon transition, this study develops a multi-dimensional hybrid carbon emission reduction estimation model (LCA-IB Method) that integrates Life Cycle Assessment (LCA) with an Improved Baseline Method. This model innovatively quantifies the carbon reduction contribution rates of individual retrofit technologies while accounting for embodied carbon in equipment and operational carbon emissions. Taking Suizhong Power Plant’s 2×800MW Russian-made thermal power units as a case study, the model was validated using 18 months of high-frequency (5-minute interval) operational data (1.2 million data points) and on-site continuous emissions monitoring system (CEMS) data. Key results show: (1) The retrofit, incorporating spray cooling, counterflow/parallel flow switching, and intelligent control technologies, achieved an annual carbon emission reduction of 192,300 tCO₂, with a 15.4% reduction in unit power generation carbon emissions (from 0.356 tCO₂/MWh to 0.302 tCO₂/MWh). The model’s prediction error was verified to be <2.9%, meeting ISO 14064’s precision requirements. (2) Technical contribution quantification revealed spray cooling (42% contribution, 80,766 tCO₂/year reduction) and counterflow/parallel flow switching (38% contribution, 73,074 tCO₂/year reduction) as core carbon reduction drivers. Spray cooling reduced summer air-cooling tower inlet temperature by 4.8±0.5℃, lowering unit coal consumption by 12.6 g/kWh; counterflow/parallel flow switching optimized cooling efficiency by 18.3% under 75% load. (3) Policy compatibility analysis with the U.S. Inflation Reduction Act (IRA) demonstrated the technology qualifies for dual subsidies: an annual carbon reduction subsidy of (6.73 million (based on 35/tCO₂) and a 30% Investment Tax Credit (ITC) for retrofit investments. In the U.S. market, the technology achieves a 4.2-year payback period, outperforming domestic U.S. retrofit solutions (average 5.8-year payback). This study provides a standardized, high-precision carbon accounting framework for thermal power air-cooling system retrofits and offers a technical-economic roadmap for global thermal power plants to achieve cost-effective low-carbon transitions.

Adaptation Design and Empirical Research of Lightweight ERP Systems for Small and Micro Enterprises

Wanyu Li — Thu, 06 Nov 2025 00:00:00 +0000

Against the global digital transformation of small and micro-enterprises (SMEs), the low adoption rate (<35% globally) and high failure rate (>60%) of traditional ERP systems have become bottlenecks for SME management upgrading. This study takes the “Qi Weijie” lightweight ERP system as the research object, conducts empirical analysis based on 15 SMEs across manufacturing, service, and retail sectors, and explores the influence mechanism of three core adaptation elements—scale adaptation, cost control, and ease of operation—on ERP application effects. A “Lightweight ERP Selection Scoring Model” with 10 quantitative indicators was constructed via Analytic Hierarchy Process (AHP), and its prediction accuracy was verified (in-sample R² = 0.85, out-of-sample R² = 0.80). Empirical results show that: (1) Scale adaptation has a significant positive impact on ERP application effect (β = 0.65, p < 0.001), with functional module customization and architectural flexibility explaining 42% of the variance; (2) Cost control presents a significant negative correlation with application effect (β =-0.58, p < 0.001), and every 10% reduction in implementation and maintenance costs leads to a 15% increase in user acceptance; (3) Ease of operation is the most influential factor (β = 0.70, p < 0.001), accounting for 49% of the variance, and intuitive interfaces and simplified processes increase system usage frequency by 2.3 times. This study enriches the theoretical framework of ERP adaptation research from the perspective of SME resource constraints and provides a scientific decision-making tool for SMEs. Validated in Beijing Mint Information Consulting Co., Ltd.’s practice, it helped 12 of 15 sample enterprises achieve a 45% average improvement in management efficiency and a 30% average reduction in operational costs.

Root Cause Tracing Algorithm and One-Click Repair Mechanism for Medical Server Failures

Zhengyang Qi — Mon, 17 Nov 2025 00:00:00 +0000

Delays in laboratory test reports at primary healthcare facilities often stem from server failures, with the lack of on-site expertise resulting in a mean time to repair (MTTR) of up to two hours. This directly hampers diagnostic efficiency and patient experience. To address this, we propose a root cause tracing algorithm and one-click repair mechanism tailored for medical scenarios. By embedding business process semantics into a fault propagation graph, we achieve zero-threshold self-healing. Methodologically, we first utilize eBPF probes to collect system metrics and align them with BPMN medical process diagrams to construct a business-aware root cause analysis model. Through random walk inference, the model identifies the top root cause within one minute. Subsequently, we encapsulate 23 HIPAA-audited repair scripts into a “one-click repair” controller using Kubernetes CRDs, achieving an average fault recovery time of 14 minutes. In a prospective cohort experiment conducted in 12 community health centers in 2024, we injected 411 faults. The results showed that the root cause localization accuracy increased from 52% to 93%, MTTR decreased from 119 minutes to 14 minutes, and the number of human interventions per fault dropped from 1.8 to 0.05. The annual maintenance cost was reduced by 60%. The bilingual usability score reached 4.7 out of 5, with no difference between English and Spanish interfaces. This study is the first to incorporate MTTR into CLIA quality indicators, providing a replicable, compliant, and language-friendly zero-threshold maintenance paradigm for resource-constrained regions.

The Construction and Empirical Study of a Brand Marketing Information Technology Maturity Model for Small and Medium-Sized Enterprises

Yanxin Zhu — Wed, 03 Dec 2025 00:00:00 +0000

Small and medium-sized enterprises (SMEs) face the dilemma of “vague stage positioning and low transformation efficiency” in brand marketing information technology (IT). 45% of these enterprises suffer from resource misallocation due to the lack of a unified evaluation standard. This paper, with the resource constraints of SMEs as the core premise, integrates the CMMI grading logic and the DTMF dimension framework to construct a four-stage maturity model (“Basic - Growth - Optimization - Excellence”) that includes six first-level indicators (such as “data collection and integration”) and 18 second-level indicators. Through stratified surveys of 192 SMEs across six major regions in China from September 2023 to March 2024 (40.1% in fast-moving consumer goods, 30.2% in catering, and 29.7% in retail), the model’s scientific validity was verified through reliability and validity tests (overall Cronbach’s α = 0.89, KMO = 0.82). The empirical results show that the average maturity score of enterprises is 48.6 points (Growth stage), with “intelligent decision-making application” (2.1 points) and “resource coordination ability” (2.3 points) being the core weaknesses. Enterprises in the fast-moving consumer goods sector, those with a size of 50-100 employees, and those in the East China region have relatively higher maturity scores. This study provides SMEs with a “self-assessment - improvement” tool and offers empirical references for governments to formulate differentiated transformation subsidy policies.