基于音频特征的工程车辆工况识别研究

高效率地使用工程车辆是工程项目管理中节约成本的有效方法，无人监管环境下工程车辆的工况识别，是实现工程车辆高效率使用的有效手段。目前以GPS等技术为核心的车辆智能管理系统未对工程车辆进行工况识别，提出一种基于GRU循环神经网络的工程车辆工况识别方法，通过对工程车辆在不同工况下产生的音频信号进行分析，从中提取Mel倒谱系数作为主要特征，构建GRU循环神经网络模型进行训练和识别。实验结果表明，该方法可以实现对工程车辆工况的有效识别。     

Construction and analysis of photovoltaic directly coupled conditions in PEM electrolyzer

The Markov model and the PEM electrolyzer system model for directly coupled photovoltaic are combined to construct an efficient and reliable working condition that fits the fluctuation characteristics of solar energy. The working condition is designed through genetic algorithm so that the average coupling efficiency of the system can reach 98.8%. Then, the durability and recovery test are conducted on the basis of the constructed conditions. It is found that the attenuation rate at the current density of 1A/cm² under the photovoltaic fluctuating condition reached 7.8mV/h, which is twice that under the constant current condition. The charge transfer impedance (R_ct) is the main factor leading to the degradation. It is proved by the recovery experiment that the increase of R_ct is related to the pollution of metal ions. After pickling to remove some metal ions, R_ct can be significantly reduced by 46.8% and 65.2%, respectively. After the durability test, the voltammetric charges under the photovoltaic fluctuating condition and the constant current condition are reduced by 48.3% and 19.1% It indicates that the photovoltaic fluctuation condition will accelerate the attenuation of the effective reaction area of MEA, which is irreversible even after pickling. It can be observed from the SEM images that the catalyst layer of MEA has more obvious peeling under the photovoltaic fluctuation condition, which is not conducive to material transmission and destroys the transmission channel of ions and electrons. This result can provide a reliable reference for the coupling design of PEM electrolyzer and renewable energy in the future.     

数控多轴正逆向制造实训基地建设探索

本文阐述了多轴正逆向制造发展趋势,基地建设的必要性,以五轴加工中心、五轴仿真模拟面板、三维扫描仪、3D打印机等为主的加工、逆向、快速成形先进前沿设备,形成多元化的先进制造实习链,满足、丰富实训内容,提升学生社会就业竞争力。     

Intrinsic Defect Limit to the Growth of Orthorhombic HfO2 and (Hf,Zr)O2 with Strong Ferroelectricity: First-Principles Insights

It is believed that promoting the fraction of ferroelectric orthorhombic phase (o-phase) through O-poor growth conditions can increase the spontaneous polarization of HfO₂ and (Hf,Zr)O₂ thin films. However, the first-principles calculations show that the growth may be limited by the easy formation of point defects in the orthorhombic and tetragonal phases of HfO₂, ZrO₂, and (Hf,Zr)O₂. Their dominant defects, O interstitial (O_i) under O-rich conditions and O vacancy (V_O) under O-poor condition, have low formation energies and quite high density (10¹⁶–10¹⁹ cm⁻³ for 800–1400 K growth temperature). Especially, O_i has negative formation energy in tetragonal HfO₂ under O-rich condition, causing non-stoichiometry and limiting the crystalline-seed formation during o-phase growth. High-density defects can cause disordering of dipole moments and increase leakage current, both diminishing the polarization. These results explain the experimental puzzle that the measured polarization is much lower than the ideal value even in O-poor thin films and highlight that controlling defects is as important as promoting the o-phase fraction for enhancing ferroelectricity. The O-intermediate condition (average of O-rich and O-poor conditions) and low growth temperature are proposed for fabricating HfO₂ and (Hf,Zr)O₂ with fewer defects, lower leakage current, and stronger ferroelectricity, which challenges the belief that O-poor condition is optimal.     

基于改进EEMD与混沌振子的配电网故障选线

提出改进的集合经验模态分解(MEEMD)和混沌振子相结合的电网故障微弱信号检测方法。首先,建立神经网络预测模型,通过神经网络对配网各线路零序电流进行短时预测,滤除故障信号中的背景信号;其次,为了检测配网发生单相接地故障后微弱的5次谐波信号,提出结合多尺度排列熵和完备集合经验模态分解(CEEMD)改进的改进的集合经验模态分解算法;处理已经滤除背景信号的故障信号,提取其第一固有模态函数作为混沌振子的输入。混沌振子对和内驱动力信号同频的外策动力信号有较高的敏感性,通过混沌振子输出的相图完成电网故障选线。     

润滑脂的溶解及铁谱分析试验研究

针对润滑脂由于流动性差而无法直接采用铁谱技术进行磨粒分析的问题,探讨适合溶解不同润滑脂的有机溶剂。选择锂基、钙基和铝基3种目前设备润滑常用的润滑脂,通过水浴加热与超声波震荡的溶解方法,对比研究8种有机溶剂对润滑脂的溶解效果;通过摩擦试验制备带有磨粒的润滑脂试样,通过旋转式铁谱仪进行制谱试验,并结合显微镜分析谱片磨粒沉积效果。结果表明:正己烷溶液、体积分数30%二甲苯和70%正己烷混合溶液适合溶解锂基润滑脂,二甲苯和四氯乙烯适合溶解钙基润滑脂和铝基润滑脂;将润滑脂试样溶液的黏度稀释到4 mm~2/s以下时,在旋转式铁谱仪上能够获得良好的制谱效果。     

Event-triggered distributed receding horizon control for tracking and formation of homogeneous multi-agent systems

This paper proposes an event-triggered distributed receding horizon control (DRHC) approach for the formation and tracking problems of homogeneous multi-agent systems. For each agent, an event-triggering condition, based on assumed predictive information of the neighbours, is derived from stability analysis. Considering the uncertain deviation between the assumed and true predictive information, we design a time-varying compatibility constraint for the individual optimization problem. In the event-triggered DRHC algorithm, each agent solves the optimization problem and communicates with its neighbours only when the event-triggering condition is satisfied, so the communication and computation burden are reduced. Moreover, guarantees for the recursive feasibility and asymptotic stability of the overall system are proved. A simulation example is provided to illustrate effectiveness of the proposed approach.     

水润滑螺旋阶梯腔艉轴承接触特性分析及结构优化

基于有限元法,考虑艉轴承材料的非线性力学特性以及各结构参数间的交互作用,对该轴承进行接触分析。采用正交试验法分析腔长、腔数、螺旋角、深浅腔比例等主要影响因素对接触应力及变形量的影响规律,采用控制变量法分析螺旋腔腔角、进/出水孔直径、螺旋腔腔深等次要因素对接触应力及变形量的影响规律,分别得到各参数的优化值。同时利用多目标优化方法,对艉轴承的沟槽直径和内衬外径进行优化,以满足低应力及轻量化的设计要求。优化后艉轴承质量降低了29%,接触应力降低了12%,在达到轻量化的目的同时,显著提高了其力学性能。     

基于历史数据聚类的火电机组工况划分

针对调峰背景下火电机组非稳态工况增多,以及常见运行工况偏离设计工况等问题,提出了基于历史运行数据聚类的工况划分模型。首先,考虑到运行数据中非稳态工况与稳态工况并存的情况,以功率作为特征变量,提出基于功率差值期望区间估计的稳态判别算法,筛选出历史数据中的非稳态工况;其次,由于稳态工况下外部边界条件变量的分布差异性,提出改进的多步K-均值聚类算法进行稳态工况的划分,并利用silhouette评价准则确定每步条件下的最佳聚类数;最后,采用某实际发电用重型燃气轮机的历史运行数据进行模型验证。通过与传统K-均值聚类算法比较,所提出的模型能够有效解决工况分类数目较少以及样本分布不均的问题。