Aiming at the problems such as low pixel,disconnected font and undetectable text line adhesion in the scanning images of vertical ionospheric data for pin printer font,an auto-matic data extraction technique based on CRNN deep learning framework is proposed,which includes four modules:image preprocessing,text detection,sequence text recognition and result layout processing.Firstly,image template matching,noise reduction and tilt correc-tion were used to preprocess the scanned images of three types of pin print vertical data with different line spacing types.Then,text detection and segmentation were performed on the preprocessed images by projection method.In the projection segmentation detection algo-rithm,vertical projection,horizontal projection and detection candidate frame correction functions were added.It can effectively deal with the cohesive text area and improve the de-tection accuracy.Finally,considering the different length of the image array,the segmenta-tion of characters is avoided,the segmented text recognition problem is transformed into a sequence learning problem,and the CRNN deep learning algorithm composed of CNN+ RNN+CTC is used for text recognition,and then the recognition results are saved into Ex-cel standardized format by coordinate fusion algorithm,so as to realize automatic data ex-traction and saving.The experimental results show that the algorithm proposed in this paper has a text detection recall rate of 97.7%,a text recognition comprehensive evaluation index F value of 97.49%for a single character recognition rate and 94.78%for a whole group of characters recognition rate,and is compared with other algorithms to verify its effectiveness.Therefore,the algorithm proposed in this paper has high practicability and can meet the ac-tual needs of engineering applications.
To optimize photovoltaic (PV) power generation efficiency and the stability of output power, the paper presents an MPPT (Maximum Power Point Tracking) technology for PV system of merchant marines. For the unstable weather, ship roll and the sea salt crystallization on solar panels under complex sea condition disturbance, the MPPT controller is difficult to design according to the accurate PV panels model. We propose a boost converter control strategy based on MFALC (Model Free Adaptive-learning Control). Firstly, a general discrete non-linear system is established according to the data of PV panels output and inputs; Secondly, a data model based on compact form dynamic linearization is carried out to design the MFALC controller; Thirdly, the pseudo partial derivative estimation algorithm is given. The proposed strategy effectively reduces the power oscillation of ship PV system and achieve MPPT rapidly under different operating conditions. The simulation results verify the effectiveness and advantages of the proposed control strategy compared with the perturbation and observation method.
We developed and validated a novel Fourier transform infrared (FTIR) method to determine the degree of molar substitution (MS) for hydroxypropyl chitosan (HPCS) using nuclear magnetic resonance (1H NMR) as a reference, and investigated the factors influencing the MS assay. Through extensive screening of integration methods for candidate bands in the FTIR spectrum of HPCS using 20 HPCS samples with degrees of acetylation (DA) ranging from 0.003 to 0.139, we found that when using band area at 2970 cm-1 as a probe integral, the MS values obtained via the 1H NMR method exhibited linear correlations (R2 > 0.98) with at least 16 integral ratios derived from their FTIR spectra. The optimal reference bands with high reliability are located at 3440 cm-1 and 1415 cm-1, with R2 exceeding 0.99 and a MS range of 0.17-1.92. The band at 2875 cm-1 is less affected by the trace moisture present in HPCS samples than the others. The results of the method validation demonstrated a mean recovery of 98.9 ± 2.8 % and an RSD below 10 %, suggesting a simple, robust, and highly accurate and precise method. This method could be extendable for the determination of the MS of insoluble HPCS derivatives and other hydroxypropylated polysaccharides.
Polyaniline (PANI) particles were synthesized using aniline (AN), dodecyl benzene sulfonic acid (DBSA), and ammonium persulfate (APS). The particles were analyzed using scanning electron microscope (SEM), X-ray diffraction (XRD), and fourier transform infrared spectroscopy (FTIR). A PANI/waterborne polyurethane composite material (PANI-WPU) was obtained by combining it with polyethylene glycol (PEG600), diphenylmethane diisocyanate (MDI), dimethylol propionic acid (DMPA), N-methyl pyrrolidone (NMP), and dibutyltin lauric acid (DBTDL). The structure was characterized by the FTIR spectrum. The mechanical characteristics of the coating film were evaluated with respect to the PANI content, as well as its water absorption, glossiness, electrochemical corrosion resistance, and acid and alkali resistance. The PANI/waterborne polyurethane film has a maximum tensile strength of 23 +/- 1$$ \pm 1 $$ MPa, an elongation of 1012%, a pencil hardness of 5H, a flexibility of 2 mm, an impact resistance of 50 cm, the water absorption of 14.66%, and the glossiness of 99.9 +/- 0.1$$ \pm 0.1 $$ at 60 degrees. When the PANI content is 0.7%, the mechanical characteristics, glossiness, and anti-corrosion performance of the composite film improve. The corrosion inhibition efficiency of the aqueous polyurethane coating film with PANI can reach 99.74%, as shown by the examination of electrochemical polarization curves and impedance spectra. The tinplate is coated with a 0.7% PANI-WPU composite material and edge sealing. This coating provides excellent protection against acid and alkali resistance, as demonstrated by its ability to withstand immersion in 10% H2SO4 and 10% NaOH solution for 90 days without any paint peeling off.
In this paper, we propose two data-driven adaptive tuning (DDAT) approaches of iterative learning control (ILC) for nonlinear non-affine systems. First, a compact-form iterative dynamic linearization (CFIDL) method is introduced to transfer the original nonlinear system into a linear data model. Then, we design an objective function for the tuning of the learning gains of a PD-type ILC law. By optimizing the designed cost function subjected to the linear data model, a CFIDL-based DDAT method is proposed, where only the real I/O data are used without requiring any mechanistic model information. Furthermore, the results are extended by introducing a partial-form iterative dynamic linearization (PFIDL) method for the purpose of utilizing more additional control information. Following the similar steps, a PFIDL-based DDAT method is developed for learning gain tuning of the PD-type ILC scheme. Both the proposed DDAT methods can help the PD-type ILC have a better robustness against to the uncertainties since they can use the real I/O data to iteratively tune the learning gains. The convergence of the DDAT-based PD-type ILC methods has been proved rigorously. The effectiveness of the two proposed DDAT-based ILC methods are further verified through simulations.
A green and effective catalytic system for rosin polymerization using a recyclable Lewis acidic DES catalyst has been developed.
围绕固液相平衡计算,综述了活度系数法、经验模型法和状态方程法的研究进展和应用.概括和分析了活度系数模型(Pitzer、ELECNRTL、Wilson、NRTL、UNIQUAC、正规溶液理论)、经验模型(λh、Apelblat、Jouyban-Acree)和状态方程模型(PR、SRK、Martin-Hou),并介绍了模型的适用范围及优缺点.固液相平衡计算模型对化工固液分离过程中的理论研究和工业应用具有重要意义.
传统的有机溶剂型防锈漆因存在严重的环境及安全问题而逐渐被水性防锈漆所替代,但后者存在干燥慢、硬度低、耐水和防腐蚀性能差等缺点.为此,对国内外对水性防锈漆大量的研究进行了归纳分析,结果表明通过采用复合成膜树脂、含磷和氟的成膜树脂、超支化成膜树脂、可交联成膜树脂、新型防锈颜料及纳米材料等能明显改善水性防锈漆的以上性能,但一些技术的采用明显增加了水性防锈漆的成本,因此降低以上技术的成本是今后研究的目标.
石墨烯纳米片是由sp2p组成的密实单层蜂窝状晶格纳米结构杂化碳原子,具有优良的热性能、优异的力学和电学性能.但由于其具有较大的表面能、较高的化学稳定性,石墨烯纳米片 自身之间存在很强的范德华力,容易发生不可逆的聚集,使其难溶于水和常用有机溶剂,限制了石墨烯的进一步研究和应用.本工作采用液相剥离法制备石墨烯,通过γ-氨丙基三乙氧基硅烷的偶联作用,制备了石墨烯白炭黑杂化材料.白炭黑的修饰改善了石墨烯的分散性问题,有效解决石墨烯的自聚集和与基体界面结合作用弱的瓶颈问题,并且能够使石墨烯稳定分散,尤其是在橡胶基体中,从而改善天然橡胶的各项物理机械性能.
针对基于孪生网络的目标跟踪算法在相似目标干扰和发生遮挡时容易丢失目标的问题,提出一种基于多注意力融合的抗遮挡目标跟踪算法(anti-occlusion target tracking based on multi-attention fusion,AOTMAF).为更好地模拟遮挡图片,引入渐进式随机遮挡模块,由易到难地随机生成遮挡块对图像进行多区域遮挡,通过人工模拟被遮挡图像的方式扩充负样本数据集,提升模型在遮挡情况下对判别性特征的提取能力.从深度、高度与宽度三个维度挖掘特征图通道信息,并通过融合空间注意力,聚合特征图上每个位置的空间依赖性,增强特征表达能力,进一步提高跟踪的鲁棒性.实验结果表明,在OTB100、VOT2018、GOT-10K公开数据集上,本研究方法在复杂场景下能有效提升跟踪精度和鲁棒性.
以胶乳海绵为基体,将海绵在气凝胶前驱体溶液中挤压浸渍,充分吸收溶胶,经老化、超临界干燥后获得具有优良隔热能力的柔性SiO2气凝胶/胶乳海绵复合材料.表征并分析了复合材料的微观结构和性能,研究了甲基三甲氧基硅烷(MTMS)的含量对隔热能力的影响,进一步探索浸渍次数对隔热及力学性能的影响.研究结果表明:当MTMS的含量(体积分数)为10％时,复合材料具有最小的导热系数.另外,在多次的浸渍处理下气凝胶布满胶乳海绵的表面和内部孔隙,成功制备出了机械性能良好、导热率达到0.038 W·(m·K)-1 的SiO2气凝胶/胶乳海绵复合隔热材料.
通过水热法在泡沫铜基体上生长MoS2前驱体,并在合成过程中通过Co的加入,优化催化剂表面形貌.Co-MoP/MoS2/Cu为无定形态,催化剂纳米颗粒呈球形,表面具有密集的刺状突起.通过一系列表征手段,证明了 Co的加入能有效改善MoP/MoS2/Cu的表面形貌,增加催化剂的比表面积,有利于催化剂与电解液的充分接触,提高催化剂的催化活性.电化学测试表明,在电流密度为10 mA·cm-2时,过电位为60 mV,具有良好的电催化析氢活性.在大电流密度下具有良好的稳定性,在300 mA·cm-2的电流密度下连续工作24 h后,性能无明显下降,能更好地满足工业化的要求.
采用基于第一性原理计算方法探讨了 Ni2B(001)晶面的电子结构和析氧反应(OER)性能.研究发现:Ni2B(001)有着优异的电导率和OER活性.同周期过渡金属掺杂可有效地调节Ni2B(001)电子特性并提高其OER活性,Cr掺杂效果尤其明显.电荷密度和d带中心分析表明,掺杂原子M周围产生了更多的电子空穴,表面原子的d带中心更靠近费米能级,电子迁移率更高,OER催化活性更强.OER四电子反应过程中的第三步,即由中间体*O至*OOH的反应为电势决定步骤,当过电势为1.500 V时,OER反应自发进行.
针对多变量受控自回归自回归滑动平均(M-CARARMA)系统,利用滤波辨识理念和递阶辨识原理,研究和提出了滤波递阶广义增广随机梯度辨识方法、滤波递阶多新息广义增广随机梯度辨识方法、滤波递阶广义增广递推梯度辨识方法、滤波递阶多新息广义增广递推梯度辨识方法、滤波递阶递推广义增广最小二乘辨识方法、滤波递阶多新息广义增广最小二乘辨识方法.这些滤波递阶广义增广辨识方法可以推广到其他有色噪声干扰下的线性和非线性多变量随机系统中.
为了规避目前金属导电墨水价格高昂、书写性能不够流畅、容易导致信号延迟的问题,在此制备了高性能的MXene和聚醚F127混合导电墨水.实验结果显示,该墨水具有流畅书写的特性,即使在界面受到应变时,其也能保持完整,确保了电信号的可靠性.此外,通过改变笔尖直径或注射直径,可以设计不同规模的电路,以适应各种应用场景.最后,成功利用这种导电墨水设计了一种电子皮肤传感器,用于监测人体的生理活动.这项研究为未来在柔性电子和生物传感器领域的应用提供了新的可能性,同时降低了成本并提高了性能.
针对目前存在的特征点SLAM(同时定位与地图构建)算法在低纹理环境下难以提取足够多的特征点、定位精度低等问题,提出一种基于结构约束的点线特征融合SLAM算法.通过LSD算法提取直线后,将直线分为平行线与垂线,以此作为约束优化特征线的三维位置与相机位姿;通过CAPE算法提取表面法向量估计环境中的曼哈顿主方向,将其作为局部地图的约束优化特征位置与相机位姿.最后,在TUM RGB-D数据集上对提出的算法进行验证.结果表明,该算法提高了在弱纹理环境中的定位精度,数据有效提高了系统的稳定性.
以某地区的配电网为研究对象,对该区域配电网的供电方式及负荷性质进行了分析,针对引起该区域电压不合格及无功不足问题的原因,建立了考虑成本效益回报和线损率的多目标无功优化模型.将模拟退火算法(SA)中的Boltzmann策略,加入到传统遗传算法(GA)中,构成改进遗传算法(SA-GA),以提升遗传算法的全局搜索能力.通过实际配电网仿真算例,验证了本模型及算法对解决该区域无功及电压问题的有效性.
为实现温和条件下甲醛或甲醇的制备,对光催化辅酶再生体系与甲醛脱氢酶(Fald-DH)耦合进行了研究,提出了以甲酸歧化方式制备甲醛、甲醇的新方法.该方法将光催化NADH再生体系与甲醛脱氢酶、醇脱氢酶耦合,不过未检测到甲醛或甲醇.通过实验对结果进行验证,考察了 FaldDH催化甲酸盐还原制备甲醛的反应的影响因素,并通过平衡反应、结合已有文献报道给出了合理的解释.结果表明:受热力学反应平衡限制,反应条件下生成的甲醛量低于Nash显色法定量甲醛的检测限,以往报道中酶催化生成的甲醛极有可能未利用NADH作为还原剂.
聚丁烯-1作为一种多晶态聚合物,在拉伸过程中晶体结构会逐渐由亚稳态的晶型Ⅱ转变为稳定的晶型Ⅰ.利用原位广角X射线衍射,研究了不同温度下聚丁烯-1在拉伸过程中的晶型转变,以及预成核过程对聚丁烯-1晶型转变的影响.结果发现:拉伸温度越高,晶型转变越慢;预成核可以加速聚丁烯-1的晶型转变;此外,拉伸温度越高,预成核对晶型转变的促进作用也越明显.
采用实验方法对比制冷剂在3根强化管和1根光滑管外环形侧的流动沸腾换热和压降特性.实验采用的3根强化换热管分别是具有螺旋微翅片的换热管(HB管)、具有凹坑结构的换热管(DIM管)和同时具有螺旋微翅片和凹坑复合结构的换热管(DIM/HB管).实验的饱和温度为6 ℃,质量流速范围是75～225 kg·(m2·s)-1,进出口干度分别保持在0.2和0.8.从流动沸腾换热实验结果可以看出,具有复合结构的DIM/HB管展示出最高的换热系数,约为光滑管的1.54～2.23倍.但同时DIM/HB管也展示出相对较高的摩擦压降,约为光滑管的1.26倍.基于文献中2种经典的蒸发换热预测关联式分别对预测模型进行修正,通过比较发现,基于Thome关联式的修正预测模型拟合效果更好,误差可达10％以内.
