NetWork
Design and experimental research on energy management strategy for fuel cell vehicles
SUN Ke;SUN Wen;BAI Shuzhan;[Objective] Under the global “Dual Carbon” goals， the transportation sector urgently needs cleaner and more efficient powertrains. Proton Exchange Membrane Fuel Cell (PEMFC) technology， featuring zero tailpipe emissions， high energy density and extended range， has become a key pathway for commercial-vehicle decarbonization. However， the slow dynamic response of PEMFC systems makes it difficult for a stand-alone fuel cell to cope with frequent start–stops and rapid load fluctuations. Therefore， fuel cell–battery hybrid architectures are widely adopted， which place higher requirements on Energy Management Strategies (EMS) to coordinate multi-energy sources efficiently. This study targets the EMS design problem for fuel cell commercial vehicles， and experimentally compares two representative strategies from an engineering perspective to provide a basis for system optimization. [Methods] A hardware-in-the-loop bench platform for a fuel cell hybrid electric vehicle was established. The platform integrates a PEMFC system， a lithium iron phosphate battery pack and a drive motor， together with a dynamic load simulation system to reproduce real-world driving resistances. A self-developed LabVIEW-based software platform combined with CAN bus communication enables real-time driving-cycle replay， data acquisition and EMS verification. On this basis， two EMSs were implemented: a deterministic rule-based strategy and a fuzzy-logic-based strategy (FLC). The rule-based strategy adopts a master–slave architecture that treats the fuel cell as the main power source and allocates power according to total vehicle demand and battery state-of-charge (SOC) thresholds. The FLC takes vehicle power demand and SOC as inputs and outputs fuel cell power， adjusting power distribution via predefined fuzzy rules. [Results] Bench tests under the same driving cycle reveal distinct characteristics for the two strategies. Both achieve similar peak and average motor power， while the FLC exhibits slightly larger power fluctuations and faster tracking of load changes， indicating better dynamic adaptability and smoother acceleration/deceleration behaviour. Under the FLC， the fuel cell operates at lower average power with more frequent start–stops， and the battery undertakes a larger share of power regulation; under the rule-based strategy the fuel cell output is more continuous and often near its rated power， and the battery SOC is maintained within a narrow band， which is beneficial to battery life. Hydrogen-consumption analysis shows that， although the FLC reduces instantaneous fuel cell usage， its higher reliance on battery charge–discharge cycling leads to higher equivalent hydrogen consumption， whereas the rule-based strategy offers shorter energy paths and better overall fuel economy. Quantitatively， the rule-based strategy achieved an actual and equivalent hydrogen consumption of 568.13 g and 628.84 g， respectively， compared with 441.41 g and 726.40 g for the FLC. [Conclusions] The two EMSs present complementary advantages. The FLC is more suitable for urban scenarios with frequent start–stops and rapidly varying loads， where dynamic response and driving smoothness are prioritised. The deterministic rule-based strategy is more appropriate for relatively stable operating conditions that emphasise range and hydrogen economy， thanks to its efficient and stable power-supply characteristics. Future work will focus on adaptive or hybrid EMS frameworks that combine rule-based logic with data-driven or intelligent optimisation methods to enhance multi-scenario applicability and overall energy-management performance of fuel cell hybrid commercial vehicles.
Magnetic-Coupling Nonlinear Internal Resonance: An Inquiry-Based Experimental Teaching Design
TIAN Wei;GAO Xiangguo;WANG Le;YANG Zhichun;[Objective] Aligned with recent developments in the research field， this work develops an exploratory experimental teaching platform specifically designed to investigate nonlinear internal resonance phenomena， with magnetic coupling effects serving as the core and essential mechanism. The primary aim of this platform is not only to enable students to systematically and comprehensively master the fundamental theoretical foundations of nonlinear vibration， but also to further enhance their abilities in independent critical thinking， experimental planning， and problem-oriented design. In addition， the platform guides students through the complete process of structural modeling， theoretical analysis， and experimental validation of internal resonance phenomena， thereby fostering a deeper understanding of the dynamic behaviors involved. [Methods] For the physical experimental setup， double cantilever beam specimens are manufactured using 3D printing technology， and an internal resonance testing platform is subsequently constructed by integrating the specimens with a vibration testing bench， laser displacement sensors， and an LMS data acquisition system. Identical permanent magnets are mounted at the free ends of the two cantilever beams， and the occurrence of internal resonance can be experimentally tuned and controlled by adjusting the stiffness characteristics of the beams. Within the theoretical modelling framework， a magnetic dipole model is employed to characterize and quantify the nonlinear magnetic forces between the magnets. On this basis， the governing equations of the magnetically coupled double-cantilever-beam system are formulated in accordance with Newton’s second law. During experimental testing， constant-amplitude sinusoidal sweep excitation is employed to the system， and its time-domain responses are recorded using an LMS data acquisition system. The acquired time-domain signals are subsequently transformed into the frequency domain via Fast Fourier Transform (FFT)， thereby facilitating clearer observation and more detailed analysis of internal resonance characteristics. [Results] Focusing specifically on the 2:1 internal resonance， the following key experimental observations are made: 1) Sweep-frequency testing reveals that the low-frequency cantilever beam， with a linear natural frequency of approximately 17.5?Hz， displays distinct amplitude peaks when excited within the 48–55?Hz range; 2) Corresponding frequency-spectrum analysis indicates that under 48?Hz excitation， the response frequency of the low-frequency beam stabilizes near 24?Hz， confirming a 2:1 internal resonance between the two structural modes; 3) When the excitation frequency is tuned to about 50?Hz， the system does not initially exhibit internal resonance under steady conditions; however， introducing an external perturbation reliably triggers a pronounced 2:1 resonant response， demonstrating the sensitive dependence of the phenomenon on initial conditions and external excitations. These results collectively illustrate the characteristic frequency-coupling behavior and activation threshold associated with 2:1 internal resonance in the magnetically coupled cantilever system. [Conclusions] Practical teaching experience and implementation results confirm that the proposed experimental platform is straightforward to assemble， employs well-standardized procedures， and demonstrates high practicality and ease of use in instructional settings. Furthermore， the platform has been shown to significantly enhance students’ grasp of fundamental nonlinear-dynamics concepts and to foster their integrated development in scientific inquiry， engineering insight， and hands-on experimental competency， thereby supporting a more inquiry-driven and applied pedagogy.
Thoughts and Discussions on the Construction of Provincial Laboratories
CHEN Jie;TAN Yue;GUO Yangxue;[Objective] Provincial laboratories represent the highest-level scientific research and innovation platforms at the provincial level. They focus on national strategic objectives and major regional industrial and technological innovation needs， integrating superior scientific and technological innovation resources， achieving breakthroughs in leading and disruptive core technologies， driving the transformation and industrialization of scientific achievements， and supporting high-quality economic and social development in the region.With the steady advancement of construction， provincial laboratories have gradually shifted from a phase of rapid quantitative growth to a phase of qualitative improvement.It is necessary to address the question of how to build and develop the laboratories well after getting them established.Further in-depth thinking and research are required on their development orientation， institutional mechanisms， operation modes， evaluation methods and other related aspects.[Methods] Based on the method of policy text analysis， this study sorts out and summarizes the characteristics of provincial laboratories from four perspectives: policy and institutional guarantee， positioning of new-type institutions， innovating in systems and mechanisms， and strengthening diversified investment. Taking the construction practices of the Shaanxi Aerospace Power Laboratory and the Shaanxi Energy Laboratory as case studies， this study summarizes their construction characteristics and effectiveness in aspects such as achievement transformation， financial support， open sharing， and diversified investment through field investigations and in-depth interviews.[Results and Conclusions]Based on the current situation of Shaanxi provincial laboratories， this study proposes paths and recommendations for their transition from rapid quantitative growth in construction to sustained qualitative improvement， focusing on the following four aspects: 1) accurately defining the positioning of provincial laboratories and refining a differentiated development path distinct from other platforms and bases. On the one hand， they should target national major strategic needs， conduct forward-looking， leading-edge and innovative research， and enhance the capacity for original innovation. On the other hand， they need to closely focus on local economic development needs， select and prioritize flagship development directions and priorities， 2) encouraging innovation， and streamlining various systems and mechanisms， including the operation of independent legal entities， cooperation among directors， transformation of research achievements and mechanisms for tolerance of mistakes and trial-and-error practices， which involve establishing scientific research entities with clear responsibilities， rights and interests， independent overall allocation， and standardized management and operation; properly handling the relationships between leading units and co-constructing units such as profit distribution and the ownership of intellectual property rights; and exploring and adopting models such as "investment before equity" and "payment after use"， 3) exploring and refining the development model of integrated convergence among universities， enterprises and research institutes， as well as the model of full-chain connectivity from basic research to industrial application， so as to promote provincial laboratories to improve their operational mechanisms characterized by multi-stakeholder collaboration， integration of all factors， full-chain connectivity and all-round support; 4)adhering to a results-oriented approach， establishing an evaluation mechanism aligned with the research characteristics and operational rules of the provincial laboratories， exploring and establishing a phased and long-term performance evaluation system through the combined approach of annual reporting and regular assessment， and setting distinct evaluation indicators for the initial construction phase and operational phase respectively.
Development and Application of an Experimental Device for Lost-Circulation Plugging in Stress-Sensitive Fractures
HU Han;FENG Yongcun;LAI Chenxi;WANG Guangyu;LI Xiaorong;[Objective] Stress-sensitive fractures may undergo continuous aperture variation under wellbore pressure fluctuations during drilling. However， many existing lost-circulation evaluation methods and material-selection criteria are established under a fixed fracture aperture assumption. Therefore， the selected plugging formulations may not match the evolving fracture width， leading to plugging-layer instability and repeated loss. To address this challenge， we have developed an experimental device that represents a coupled wellbore-formation-fracture plugging system. [Methods] The device consists of a drilling-fluid circulation system， a temperature-control system， a plugging-slurry preparation system， a dynamic fracture-aperture regulation system， and a data acquisition system. It was developed to simulate lost circulation in stress-sensitive fractures with pressure-driven aperture evolution which enables independent control of fracture closure pressure， fracture-inlet pressure， temperature， and different fracture geometries. Pressure， temperature， and fracture deformation can be recorded synchronously， and loss volume can be quantified. After each test， the fracture module can be disassembled for direct observation of plugging-layer distribution and local bridging morphology. Two plugging formula systems were compared using the developed device. A conventional bridging system composed of rigid bridging and filling materials， and an elastic particle enhanced system using recycle tire rubber particles were investigated to improve deformation adaptability under dynamic conditions. The effects of fracture closure pressure， total lost circulation materials (LCM) concentration， plugging-material system， and composite plugging formula performance were systematically tested. [Results] The comparative tests indicate that increasing fracture closure pressure can improve plugging stability against dynamic aperture growth， but the improvement is only 7.5% when the closure pressure increases from 1 MPa to 3 MPa， indicating that fracture closure pressure alone has limited ability to improve dynamic plugging stability. Total LCM concentration exhibits an effective operating window. When the concentration is too low， persistent leakage occurs because a continuous and stable bridging skeleton cannot be established. When the concentration is too high， premature entrance plugging may occur， which increases the apparent pressure-bearing capacity but may result in operational risks in circulation control. The effect of elastic particles is strongly dependent on fracture aperture. In 1 mm fractures， the elastic particle enhanced system failed to form a stable pressure-bearing plugging layer， indicating that small-aperture fractures rely primarily on the rapid establishment of a stiff bridging skeleton. In 2 mm fractures， the elastic particle enhanced system reduced fluid loss from 188 mL to 133 mL under comparable conditions. This result demonstrates that elastic particles are more effective in larger fractures， where deformable filling and contact reconfiguration improve loss control. After further optimization of the rubber particle size combination， the fluid loss was reduced from 133 mL to 98 mL， while the pressure-bearing capacity remained within 4.5-4.8 MPa. This indicates that particle-size recombination mainly improves loss control and plugging-layer densification rather than markedly increasing the ultimate pressure-bearing strength. [Conclusions] The present study establishes a dedicated experimental basis for evaluating pressure-bearing plugging in stress-sensitive fractures. The coupled wellbore-formation-fracture plugging device makes it possible to link macroscopic plugging performance with the evolution of plugging-layer morphology during dynamic fracture deformation. The experimental results provide a reliable basis for dynamic plugging evaluation， mechanism-oriented formulation design， and the development of pressure-bearing plugging strategies for fractured formations.
Experimental Platform for Performance Testing of Exoskeleton Joint Motors based on ZYNQ clip
ZHU Chunxiang;DUAN Yuxiang;ZHOU Xingkai;QI Jialong;WANG Binrui;[Objective] Exoskeleton joint motors used in exoskeleton robots are required to operate under complex working conditions such as variable loads， frequent start–stop motions， and external disturbances， which place high demands on their dynamic control performance and response capability. However， existing motor performance test systems are mostly based on discrete hardware architectures and external data acquisition devices， resulting in low integration levels， limited real-time performance， and difficulties in multi-parameter synchronous measurement. To address these problems， this paper aims to design a highly integrated and real-time experimental platform for performance testing of exoskeleton joint motors based on a ZYNQ heterogeneous system， enabling accurate evaluation of both steady-state and dynamic performance indicators. [Methods] An exoskeleton joint motor performance test platform based on the ZYNQ-7010 system-on-chip is developed using a hardware–software co-design approach. The ARM processing system is responsible for test management， parameter configuration， communication， and data processing， while the programmable logic implements high-speed synchronous acquisition of torque and speed frequency signals， analog signal sampling， and dynamic load signal generation. A drag-type experimental structure is constructed by coaxially coupling the tested exoskeleton joint motor and the load motor through a torque sensor， allowing simulation of typical operating conditions encountered in exoskeleton applications. An equal-precision frequency measurement algorithm is implemented in programmable logic to ensure accurate frequency acquisition over a wide dynamic range. Meanwhile， a direct digital synthesis–based method is adopted to generate controllable and repeatable dynamic loading signals. Based on relevant national standards， experimental procedures are established to evaluate torque ripple coefficient， speed ripple coefficient， control accuracy， and step response time.[Results]The experimental results demonstrate that:1）An embedded measurement and control hardware system based on the ZYNQ-7010 is designed and implemented， integrating high-precision frequency signal acquisition interfaces with a measurement range from 1 Hz to 1 MHz， analog signal acquisition modules， and dynamic load simulation output modules;2）A drag-type physical test platform for exoskeleton joint motors is constructed， which can simulate various typical load conditions encountered during actual motion processes and supports coordinated multi-parameter testing;3）Based on a hardware–software co-design approach and in accordance with relevant national and industry standards for exoskeleton joint motor performance testing， platform-level implementation and experimental verification are carried out for key performance indicators， including torque ripple coefficient， speed ripple coefficient， control accuracy， and step response time， with data update periods ranging from 1 ms to 500 ms.
Research on visual distance measurement system for overhead transmission lines based on improved SGBM algorithm
Liu Xiuting;Gao Feng;[Objective] The overhead transmission lines are the main channels for power transmission and a critical guarantee for residents' daily lives and industrial production， widely distributed across major roads in cities， districts， counties， and towns nationwide. The municipal engineering construction was an important initiative to promote urban development and improve people's quality of life， characterized by a large number of engineering projects， large scales， and the frequent employment of engineering construction machinery. For the purpose of preventing the external damage to overhead power transmission lines and ensure the operational safety of engineering construction machinery and relevant workers， it is very necessary to develop a binocular vision ranging-based safety warning system for overhead transmission lines targeting construction machinery. [Methods] In hardware modules， the developed safety warning system mainly consists of a perception layer， a computation layer， and an application layer. In which， the perception layer includes the CMOS binocular cameras for real-time cable image acquisition and the Raspberry Pi modules for low-latency local area network (LAN) image transmission; and the computation layer includes mainly a host computer for in-depth image processing efficiently; and the application layer includes mainly the voice chips for prompts， the electromagnetic buzzer control， the LED beads， and a power supply module. In addition， the key software modules were selected seriously and improved to improve the operational performance of safety waring system effectively in realistic conditions. For example， the Zhang's calibration method recognized by numerous scholars was taken into consideration for 3D calibration effectively， and the Bouguet rectification method considered widely was also employed for the epipolar rectification and the improvement of image distortion effectively. Specifically， an improved semi global block matching (SGBM) algorithm was proposed and used to enhance the stereo matching performance. In which， the adaptive windows， the noise tolerance variables， and the RGB channel absolute color difference (ACD) costs were introduced simultaneous into the traditional CT (Census Transform) method to obtain an improved cost calculation approach; and then， a four-path cost aggregation strategy was carried out for the initial matching cost aggregation; next， the Winner-Take-All (WTA) algorithm was used to solve the initial disparity of the target image; finally， the obtained initial disparity was optimized to solve the optimal disparity of target image by means of the median filtering algorithm.[Results] The testing results show that compared with the traditional SGBM algorithm， the image processing time of improved SGBM algorithm (3.528ms) increases by 18.26%， but its overall error rate (2.79%) and the non-occlusion error rate (2.16%) decrease by 95.08% and 95.04% respectively; meanwhile， the relative error between the detected values of the binocular vision-based overhead transmission line ranging system and the actual distances was consistently maintained within the range of 0.691% to 2.482%.[Conclusions] In a word， the binocular vision-based overhead transmission line ranging system developed in this research exhibits the high measurement accuracy， the high detection efficiency， and the good operational stability in realistic condition， which can ensure the working operation safety of engineering construction machinery effectively and prevent the external damage to overhead power transmission lines obviously.
Development and Application of an Integrated GTAW and GMAW Arc Additive Manufacturing Experimental Platform
LI Yongcun;LIU Zeguo;ZHANG Jiabao;WANG Yong;[Objective] Wire Arc Additive Manufacturing (WAAM) acts as a pivotal technology within the "Made in China 2025" strategy， distinguished by its superior deposition efficiency and material utilization compared to laser or electron beam-based additive manufacturing methods. Despite its industrial potential for manufacturing medium-to-large metal components， the widespread adoption of WAAM in academic research and small-to-medium enterprises is currently constrained by the prohibitive costs and closed-source architectures of commercial systems. Conversely， existing low-cost open-source platforms often fail to meet necessary standards for motion control precision， forming stability， and system extensibility， particularly regarding the integration of multiple welding processes. To bridge this gap， this study aims to design and construct a cost-effective， desktop-level experimental platform that integrates both Gas Tungsten Arc Welding (GTAW) and Gas Metal Arc Welding (GMAW) processes using an open-source control architecture， thereby providing a robust tool for process validation and engineering education.[Methods] The experimental platform was developed using a modular， open-source architecture featuring a moving gantry three-axis mechanical structure. To optimize cost and spatial efficiency， both GTAW and GMAW torches were integrated onto the Z-axis. High-torque 86×80 stepper motors (4.5 N·m) driven by high-subdivision drivers and ball screw transmission were selected to ensure positioning precision and structural rigidity. The control system adopts a master-slave configuration where a PC generates G-code trajectories and an Arduino Uno running Grbl firmware executes real-time motion control， synchronizing arc ignition/extinction (M8/M9 commands) and gas supply via relay modules. To mitigate oxidation in local shielding， a novel dual-path gas system was designed， combining standard torch delivery with a bottom micro-pore supply embedded in the fixture. Single-bead deposition experiments using Q345B steel substrates and ER50-6 wire were conducted to systematically investigate the effects of travel speed (0.24–0.48 m/min) and wire feed speed (5-7 m/min) on bead geometry and forming quality.[Results] Experimental analysis revealed a distinct linear correlation between process parameters and bead geometry. Specifically， increasing the wire feed speed from 5.0 to 7.0 m/min significantly expanded the bead width from 3.889 mm to 6.115 mm and increased reinforcement from 1.252 mm to 4.029 mm， driven by the higher deposition rate. However， excessive wire feed speed (7.0 m/min) compromised stability， causing severe spatter and undercutting， while excessive travel speed led to snake-like defects and discontinuity. The optimal parameter combination was identified as a travel speed of 0.36 m/min and a wire feed speed of 6 m/min. This setting achieved a dilution rate between 5% and 15% and facilitated the fabrication of complex structures like single-bead multi-layer walls without macroscopic defects. Microstructural analysis confirmed a matrix of proeutectoid ferrite， acicular ferrite， bainite， and pearlite. Notably， regional variations were observed: the bottom zone featured coarse grains due to substrate quenching; the middle zone formed interlaced acicular ferrite under moderate cooling; and the top zone contained side-plate ferrite and pearlite induced by solute enrichment.[Conclusions] The developed platform successfully integrates GTAW and GMAW processes within a low-cost， open-source framework， achieving a balance between cost-effectiveness and high control precision. By enabling the fabrication of well-formed metal components with sound microstructures， the platform proves its viability as a versatile and economical solution for WAAM process exploration， teaching demonstrations， and fundamental research in universities and research institutions.
Research on Safety Risk Prevention and Emergency Capacity Building in University Laboratories
Lu Donglin;Wu Xuezheng;Yang Ye;Chen Renshou;[Objective] The prevention and control of safety risks and emergency response capabilities are the foundation of laboratory safety construction in universities. The article intends to analyze the national regulations and accident investigation reports on emergency management of university laboratories， and summarize four issues in laboratory risk prevention and emergency capacity building. On the basis of practice， measures are proposed to strengthen the construction of safety risk prevention and emergency response capabilities in university laboratories from five dimensions: evaluation， guarantee， capability， contingency plan， and education. This provides a reference for other universities to further improve laboratory safety construction. [Methods]Through literature review， investigation and research methods， the article believes that there are still four problems in the safety of university laboratories， including weak risk identification ability， lack of operational standards for risk assessment， insufficient accident risk analysis ability， and inadequate implementation of emergency plans， which are difficult to meet the safety needs of the rapid development of university laboratory construction. To further clarify the solution path， the article used expert consultation method and convened domestic experts through a special meeting to further clarify the research path from laboratory risk prevention and emergency capacity building.[Results] Based on the relevant regulations on emergency management in university laboratories and research conducted by universities， this article summarizes three commonalities between safety risk prevention and emergency capacity building. Firstly， both are in the lifecycle of laboratory risks， which is an important measure to effectively prevent and control accidents. Secondly， in order to achieve the "forward shift" of emergency capacity building， managers must expand the sudden response to accidents to the "risk prevention and control" link. Thirdly， safety risk prevention and control is a principle requirement for ensuring laboratory construction， while emergency capacity building is a necessary measure to reduce laboratory accident losses. The article proposes five measures to address the issues of laboratory risk prevention and emergency response capabilities. Firstly， establish risk assessment standards and implement risk classification and control measures; Secondly， establish an accident case database to enhance the ability to identify hazards; Thirdly， establish a funding guarantee mechanism to enhance the level of protective material construction; Fourthly， improve emergency response plans and clarify the operational standards for drills; Fifth， strengthen the construction of emergency science popularization and improve the incentive mechanism for emergency science popularization. [Conclusions] Driven by policies， universities have gradually clarified their responsibilities for laboratory safety management， shifting from an accident driven management model to a focus on prevention. Through risk analysis and hazard identification， some hidden dangers have been eliminated， and diversified emergency drills have been carried out. By further strengthening the construction through five dimensions of evaluation， guarantee， capability， contingency plan， and education， the laboratory can achieve the goals of implementing risk classification and control measures， improving the ability to identify hazards， enhancing the level of protective material construction， clarifying exercise operation standards， and improving emergency science popularization incentive mechanisms. This can effectively enhance the safety construction of university laboratories. At present， measures related to laboratory safety construction have attracted the attention of education authorities.
Establishing an FMEA–HACCP-based access management system for university laboratory projects
SUN Dun;WU Weilin;LI Xiangguo;WANG Xiuhua;PIAO Jin;[Objective] University laboratories serve as core venues for scientific research， teaching， and talent cultivation. Their safety management directly impacts research continuity， teaching stability， and personnel safety. However， laboratory accidents frequently occur in Chinese universities， primarily during the use and storage of hazardous chemicals， as well as during equipment operation. The root causes of these accidents lie in fragmented risk identification， lagging control measures， formalistic access mechanisms， and the inherent limitations of a “post-incident remediation” model. This study establishes a scientific and operational access management system for experimental projects， targeting university research initiatives. The core objective is to “prevent experimental risks at their source and ensure the safe execution of projects.” This transforms laboratory safety management from a reactive response to proactive prevention， thereby guaranteeing secure experimental implementation. [Methods] First， by reverse engineering the root causes of university laboratory incidents and integrating core requirements from the University Laboratory Safety Inspection Checklist (2025 Edition)， we established five-dimensional prerequisites for experimental project access: personnel， equipment， materials， methodology， and environment. Building upon this foundation， we innovatively fused failure mode and effects analysis (FMEA) with hazard analysis and critical control point (HACCP) theory (fusion termed FMEA–HACCP) to construct a “risk identification–quantitative assessment–critical control point (CCP) determination–access verification” coordinated system， where the three-dimensional quantitative model of FMEA (risk priority number (RPN) = severity × occurrence × detectability) quantifies risk grading， whereas HACCP theory identifies CCPs， establishing control plans with key thresholds， monitoring protocols， and dynamic corrective actions to create a closed-loop management mechanism. [Results] After two years of practical implementation in the Agricultural Experiment Teaching Center， this access management system demonstrated significant outcomes: among 62 initial risk points， the high-risk points (RPN ≥ 301) decreased from 18 to 0; the average RPN across the entire process dropped from 286 to 123， representing a 57.0% reduction; three newly identified potential risk points， including “mixed storage of experimental waste liquids” and “operation of new instruments，” were controlled at low-risk levels (RPN ≤ 85) through early intervention， with no safety incidents occurring throughout the process. The compliance rate for CCPs rose from 68.0% to 98.5%， with hazardous chemical accounting， instrument calibration， and firefighting equipment achieving 100% compliance. Minor deviation frequencies decreased from an average of 5 incidents per month to 0.3 incidents， while the corrective response time shortened from 30 to 8 min. At the personnel level， undergraduate safety exam pass rates rose from 75.0% to 98.2%， graduate operational assessment pass rates reached 97.8%， and noncompliant operation incidents decreased by 91.7%. The management model successfully transitioned from “post-incident rectification” to “pre-emptive prevention–process control–post-verification，” thereby reducing the hazard rectification cycle from 72 to 24 h. [Conclusions] Overall， this study successfully establishes an FMEA–HACCP-based access management system for university laboratory projects by leveraging the comprehensive quantitative assessment of FMEA with precise control from HACCP. Centered on the five-dimensional prerequisites: “personnel， equipment， materials， methods， and environment，” this system enables comprehensive systemic risk prevention and control. Its operability， traceability， and scalability are fully validated through practice. The system effectively addresses the shortcomings of traditional management models， significantly reducing experimental risks while enhancing the safety literacy of relevant personnel. Ultimately， it provides a scientific paradigm for university laboratory safety management that can be extended to various laboratory types.
AI for empowering the construction and development of university laboratories in the new era
REN Guanghui;[Objective] As a “source of innovation” and “primary base” for cultivating outstanding talents， university laboratories are directly linked to the quality of higher education and the enhancement of national innovation capacity. With the rapid advancement of artificial intelligence (AI) technology， the intelligent upgrading of university laboratories has become a significant concern. Traditional laboratory models are hampered by multiple bottlenecks: inefficient allocation of experimental resources， with high idle rates of large-scale instruments and equipment and a lack of interdisciplinary sharing mechanisms; outdated laboratory management modes characterized by high manual operation and maintenance costs and weak early warning systems for safety hazards; limited experimental teaching functions that rely on fixed processes， making it difficult to cultivate innovative talents; and limited scientific research collaboration capabilities， including a lack of intelligent platforms that facilitate multi-team and cross-regional cooperation. However， the rapid development of AI technology has provided new possibilities for addressing these challenges. This study aims to systematically explore the core logic， practical paths， and value effects of using AI to update university laboratories and construct a theoretical framework for the deep integration of AI and university laboratory development. It also identifies potential application scenarios for AI technologies in laboratory resource management， experimental teaching reforms， and scientific research. The study proposes laboratory construction standards and development strategies that integrate AI， providing theoretical support and practical references for universities to transform their laboratories toward “intelligence， openness， and collaboration.” [Methods] To effectively address the challenges facing today’s university laboratories， including talent cultivation， research demands， and management efficiency bottlenecks， this study combines bibliometric analysis， theoretical construction， and case-based empirical research. [Results] The results demonstrate that AI’s three core features of perceptual， cognitive， and decision-making intelligence can be effectively applied in key scenarios such as procuring experimental equipment， constructing smart experimental platforms， and managing laboratories. For example， AI technology could optimize hardware management in laboratories and enable the creation of virtual–reality integrated experimental environments through knowledge graphs and digital twins. Additionally， it could promote scientific research collaboration platforms that can transcend disciplinary boundaries， improve research efficiency， and gradually form a positive cycle of “technology optimizing management—management feeding back into teaching—teaching and research collaborating.” The results also clarify the major challenges in upgrading laboratories， including high technical integration barriers， substantial costs for system upgrades， and shortages of interdisciplinary professional talents. To this end， the study proposes building a collaborative promotion mechanism of “demand-driven—technology adaptation—institutional guarantee，” strengthening talent cultivation， and improving ethical norms to advance laboratories toward higher goals of being “intelligent， open， green， and sustainable.” [Conclusions] This study systematically addresses the critical question of “how AI empowers the construction and development of university laboratories.” The results indicate that AI can effectively support the transformation and upgrading of laboratory operation models from “human-driven” to “intelligence-driven.” This conclusion aligns with the strategic orientation of digital transformation in higher education and provides a Chinese solution for global intelligent laboratory construction. It also holds significant importance for promoting the transition of university laboratories from supporters to leaders.