MEMS陀螺零偏温度补偿方法

MEMS陀螺零偏是惯性导航系统的主要误差源之一,而外界环境温度变化对MEMS陀螺零偏具有重要影响。针对上述情况,提出基于小波阈值去噪与RBF神经网络相结合的零偏温度补偿方法。通过设计好的实验方案采集与温度相关的MEMS陀螺输出数据,并采用不同的温度补偿方法进行零偏温度补偿。实验结果表明,与原始输出、多项式拟合及RBF神经网络相比,基于去噪RBF神经网络的零偏温度补偿方法精度更高,适应性更强,有效地提高了MEMS陀螺输出精度,进而提高惯性导航系统精度。     

MEMS陀螺零偏误差补偿方法研究

在高动态、恶劣温度环境下,MEMS陀螺仪零偏不仅受温度变化的影响,同时还受线、角运动等影响,其真实误差是所有因素耦合的结果。针对MEMS陀螺零偏温度和转速非线性耦合误差补偿问题,根据径向基(RBF)神经网络原理,提出了一种新的零偏误差补偿方法,并利用一种隐式结构MIMU对补偿效果进行比较,验证了采用RBF神经网络对低精度MEMS陀螺零偏误差补偿的有效性。     

利用测温电路线性补偿MEMS加速度计零偏温漂

为提高MEMS加速度计温度稳定性,针对MEMS加速度计零偏温度漂移问题,提出了一种基于闭环点位置控制的零偏温漂抑制方法,该方法利用测温电路生成线性控制电压,在不同温度下将闭环点位置控制于零反馈位置附近,此时闭环加速度计的零偏近似为0,在全温范围内的零偏漂移量可显著降低。对闭环MEMS加速度计进行温度循环试验,试验结果表明,加入线性控制电压后闭环加速度计的零偏温漂下降到0.1 mg/℃以内,与不加线性控制电压相比提高了一个数量级以上。该方法实现简单,可操作性强,无需额外硬件开销,利于工程化实现。     

基于改进SPI的多量程MEMS虚拟陀螺研究北大核心CSCD

文中设计了多量程MEMS虚拟陀螺,解决了现阶段微机电系统(MEMS)虚拟陀螺不能兼容载体对陀螺仪精度和量程的问题。提出一种改进SPI的新型数据采集方式,在改善数据读取速率的同时增强了数据实时性。在静态测试中,该阵列经过滤波融合后,1σ标准差降低了76.8%,零偏不稳定性降低了89.1%;在模拟动态测试中,通过选取最佳量程数据进行滤波融合与姿态解算,该阵列最终姿态角偏离程度最小。实验结果表明多量程MEMS虚拟陀螺融合效果与降噪能力都优于单一量程虚拟陀螺。     

基于改进径向基函数神经网络的激光陀螺温度补偿

传统的径向基神经网络(RBFNN)在激光陀螺零偏的温度补偿过程中会由于随机选取中心不合适而导致算法效率降低和数值病态,故本文提出了一种基于Kohonen网络和正交最小二乘(OLS)算法的RBFNN温度补偿方法。介绍了该方法的原理及建模步骤,设计了常温和变温环境下激光陀螺的数据采集试验及其温度补偿试验。由于结合了Kohonen网络的模式分类能力和OLS的优化选择能力,该方法可以快速、准确地辨识出受温度影响的激光陀螺零偏。利用逐步回归法、RBFNN法及其改进方法对多种温变环境影响的激光陀螺零偏进行了辨识与补偿试验,试验结果表明,在常温环境下,三者的辨识能力相当;随着温变速率的上升,改进RBFNN法不仅节省了时间,其补偿后的零偏也均小于5×10-4(°)/h(1σ),提高精度均能达86%以上。得到的结果表明改进RBFNN法提高了辨识精度且稳定、有效,适用于多种温度变化环境下激光陀螺零偏的温度补偿。     

基于刚度轴偏角预估机制的 MEMS 多环谐振陀螺全闭环控制方法

针对微机电系统多环谐振陀螺正交闭环回路存在控制误差问题,提出一种基于刚度轴偏角预估机制的多环陀螺全闭环控制方法。该方法通过对微机电系统多环谐振陀螺刚度轴偏角预估,实现正交闭环回路参数自动优化调整。同时,提出了基于刚度轴偏角预估机制的全数字化闭环控制方法,实现微机电系统多环谐振陀螺的驱动、检测、正交、模态匹配环路的全闭环控制。该方法可提升正交闭环回路信噪比,增强陀螺正交漂移的抑制能力,降低陀螺零偏输出,改善陀螺的零偏不稳定性。实验结果表明,采用本文提出的基于刚度轴偏角预估机制的全闭环控制方法后,微机电系统多环谐振陀螺的零偏输出由0.201°/s降低为0.021 3°/s,零偏不稳定性由39.42°/h降低为1.237°/h,分别降低了9.44倍和31.86倍,验证了该方法对提升微机电系统多环谐振陀螺仪性能的有效性。     

一种陀螺与罗盘组合导航系统的故障检测方法

为了提高MEMS陀螺和磁罗盘组合导航系统的可靠性,减少各种干扰和故障对导航精度的影响,提出了一种及时、精确、有效的故障检测方法。该方法首先对正常工作时的MEMS陀螺和磁罗盘信号中的各种噪声进行分析;然后针对组合系统中的常见故障,从陀螺和磁罗盘航向信息中提取出短时间内两者航向变化量差值的均值、均值的一阶差分量及二阶差分量作为故障检测量,用以区分识别不同的故障源;最后使用预警模式,来减小故障检测过程中的延迟影响,提高故障检测的及时性。实验结果表明,此方法可以及时、准确地区分识别出组合系统中零偏、速率斜坡故障以及两种磁干扰,有效地降低了干扰和故障的引入误差。     

光纤惯导温度效应误差的高精度综合辨识与补偿

为了提高光纤惯导系统在温变环境中的使用精度,研究了惯性仪表温度效应误差的高精度综合辨识与补偿。首先利用全温范围下系统级标定技术建立了高精度标定参数基准并获得了陀螺和加速度计标度因数与加速度计零偏的温度系数。之后借助工作温度下仪表坐标系与基准温度下本体坐标系间的方向转换,确定了仪表安装误差与本体坐标系随温度的变化关系并修正了温度引起的坐标系间指向误差。在此基础上,给出了陀螺零偏温度漂移的分离与补偿方法。常温静态与变温动态导航试验结果表明,前者定位误差由2 093降至442 m,后者5 h位置精度则提高了接近5倍,显著增强了惯导系统的温度适应性。     

基于带温箱单轴转台的光纤陀螺自动测试系统设计

光纤陀螺的标度因数和零偏稳定性是影响其测量精度的重要指标。光纤陀螺对温度较为敏感,因此需要对其进行使用温度范围内的高低温测试,以全面评价光纤陀螺的性能。光纤陀螺测试时,需要控制转台和温箱,同时保存和处理光纤陀螺的输出数据,操作烦琐,容易出错。带温箱单轴转台的光纤陀螺自动测试系统能够按输入转速点、温度点进行速率测试、位置测试以及高低温测试,实现带温箱单轴转台的自动控制,并生成测试报告。     

闭环光纤陀螺零偏与标度因数的综合补偿

温度特性和非线性是影响光纤陀螺精度的重要因素,为研究闭环光纤陀螺的复合模型及补偿方法,在组建的测试系统下,在全温范围内各速率点处分别测试闭环光纤陀螺仪的标度因数和零偏。根据所测结果,分别建立与温度、速率相关的零偏和标度因数非线性模型,采用多项式回归分析的方法确定模型的参数。通过实测验证:建立的模型能够较好地反应光纤陀螺的温度与标度因数非线性特性,采用该模型对IFOG进行综合补偿后,其精度有了较大的提高。相比于传统方法,该方法简单,可靠,经济且易于实现。     

Influence of Plasma Treatments on the Frictional Performance of Rubbers

The frictional performance of several rubbers after pulsed-DC plasma treatments has been examined. In all cases, the treated rubbers showed better performance than the corresponding untreated ones. Stronger treatments, in terms of longer process time and/or higher substrate bias voltage, led to larger reductions of coefficient of friction and wear. The addition of hydrogen to the argon plasma did not show any additional positive effect. Nevertheless, different degrees of improvement were observed for different rubbers. In fact, the energy consumed during the tribotest scales with the maximum working temperature of the rubbers, indicating that the plasma treatment is more effective in the case of more sensitive rubbers.     

硅微谐振式加速度计的实现及性能测试

为了提高硅微谐振式加速度计性能,从一种基于DDSOG(Deep Dry Silicon on Glass)工艺的硅微谐振式加速度计样机入手,介绍了加速度计的结构、加工方法和接口电路。该谐振式加速度计结构包括敏感质量块、谐振器和微杠杆3部分,采用差动结构来减小共模误差的影响。接口电路中采用了自动增益控制电路来稳定谐振器的振幅,成功实现了谐振器的闭环自激振荡和频率检测。分析了谐振式加速度计频率输出与加速度输入的关系,测试了硅微谐振式加速度计样机性能,结果为量程±50g,标度因数143 Hz/g,零偏稳定性1.2 mg,零偏重复性0.88 mg,阈值170μg。文章最后提出,DDSOG工艺中采用的玻璃材料和硅材料温度系数不同,影响了加速度计的温度特性,因此需要进步一改进加工工艺。     

硅微谐振式加速度计温度补偿方法研究综述

硅微谐振式加速度计具有体积小、成本低、动态范围宽、高精度准数字频率信号输出等优势,但零偏、标度因数等关键性能指标受到了温度等因素的制约,尚不能满足高精度导航制导的高性能要求。因此,本文在简要介绍硅微谐振式加速度计温度特性及温度误差来源的基础上,综述了近年来国内外学者针对硅微谐振式加速度计进行的温度补偿方面的研究,包括无源温度补偿技术、有源温度补偿技术;介绍了无源温度补偿技术与有源温度补偿技术常用的方法和最新的研究成果;分析总结了各种方法的优缺点,提出探索更加精确的测温手段、热隔离效果更好的隔离装置以设计一种低功耗、预热时间短、控制稳定性强的微烘箱系统进行器件层的加热控温,以及寻求无源补偿技术与有源补偿技术相结合的更多可能性以获得温度补偿最优组合是后续温度补偿工作的重点研究方向。     

Study on the effect of measuring methods on incident photon-to-electron conversion efficiency of dye-sensitized solar cells by home-made setup

An experimental setup is built for the measurement of monochromatic incident photon-to-electron conversion efficiency (IPCE) of solar cells. With this setup, three kinds of IPCE measuring methods as well as the convenient switching between them are achieved. The setup can also measure the response time and waveform of the short-circuit current of solar cell. Using this setup, IPCE results of dye-sensitized solar cells (DSCs) are determined and compared under different illumination conditions with each method. It is found that the IPCE values measured by AC method involving the lock-in technique are sincerely influenced by modulation frequency and bias illumination. Measurements of the response time and waveform of short-circuit current have revealed that this effect can be explained by the slow response of DSCs. To get accurate IPCE values by this method, the measurement should be carried out with a low modulation frequency and under bias illumination. The IPCE values measured by DC method under the bias light illumination will be disturbed since the short-circuit current increased with time continuously due to the temperature rise of DSC. Therefore, temperature control of DSC is considered necessary for IPCE measurement especially in DC method with bias light illumination. Additionally, high bias light intensity (>2?sun) is found to decrease the IPCE values due to the ion transport limitation of the electrolyte.     

N-IGBT 正偏压温度不稳定性阈值电压综合退化模型

N型绝缘栅双极型晶体管(N-IGBT)凭借其优良性能广泛应用于现代工业各个领域,预测特定条件下器件退化情况对提高N-IGBT可靠性具有重要意义。然而,随着N-IGBT制程的降低,因正偏压温度不稳定性(PBTI)造成的栅氧化层退化进一步加剧,退化宏观表现为器件剩余有用寿命(RUL)的降低和阈值电压的改变。基于经典Power Law模型和Arrhenius模型,以退化时间为切入点,提出了相对精度更高的三阶段Power Law-Arrhenius综合退化模型;通过加速退化实验模拟了正偏压温度不稳定性对N-IGBT的退化作用,并在退化后对反映功率器件剩余有用寿命的特征参数阈值电压进行测量;基于遗传优化算法和加速退化实验数据对综合退化模型参数进行优化拟合,确定了N-IGBT综合退化模型的一般数学表达形式,得出其精度在85%以上,并高于传统的Power Law模型精度。     

硅微振梁式加速度计的温度检测及闭环控制

为实现硅微振梁式加速度计系统芯片级温度测量及系统闭环,本文针对系统的非惯性结构部分提出了微机电系统（MEMS）结构温度的芯片级测量和闭环控制优化方法。与以温控罩的温度作为参考温度的方法不同,该方法提出了供芯片级温度测量的MEMS结构、工艺及配套电路,通过直接测量MEMS结构的温度完成实时补偿,从而提高了测量精度。该方法在闭环控制的前置电路中应用了二极管电容解调电路,与前期使用的跨阻或者跨导方案相比,对器件的要求从pA级降至nA级。运用时域方法求得二极管电路方案的解析解,提出参数优化设计方法,保证了电容测量输入与输出间的线性关系。最后,采用二阶最优模型对闭环控制的后置电路进行参数优化,控制了上电时间。配合硅微振梁式加速度计原理样机进行了实验。实验结果表明,温度补偿后的零偏稳定性为52.0 μg,标度因子稳定性为16.0×10^-6,分辨率为34.9 μg。这些结果验证了本文理论的可行性。     

Temperature propagation through a mercury vapour calibration source and assessment of possible analytical biases caused by measurement of temperature variations

A combination of experimental data, and validated modelling studies, has shown that external temperature variations propagate very quickly within mercury vapour calibration apparatus, and thus any internal temperature inhomogeneities may be largely neglected. However the study has shown that the response time of the temperature measuring device used to make readings within the mercury vapour calibration apparatus can impose a bias as a result of the device’s inability to respond quickly to changes in temperature within the mercury vapour calibration apparatus. The size of this possible bias has been quantified and strategies for its elimination proffered.     

Arc Evaporation of Ti–Al–Ta–N Coatings: The Effect of Bias Voltage and Ta on High-temperature Tribological Properties

Recently, titanium aluminium tantalum nitride (Ti–Al–Ta–N) coatings have been shown to exhibit beneficial properties for cutting applications. However, the reason for the improved behaviour of these coatings in comparison to unalloyed Ti–Al–N is not yet clear. Here, we report on the tribological mechanisms present in the temperature range between 25 and 900 °C for this coating system, and in particular on the effect of the bias voltage during deposition on the tribological response. Based on these results, we provide an explanation for the improved performance of Ta-alloyed coatings. An industrial-scale cathodic arc evaporation facility was used to deposit the coatings from powder metallurgically produced Ti₄₀Al₆₀ and Ti₃₈Al₅₇Ta₅ targets at bias voltages ranging from −40 to −160 V. X-ray diffraction experiments displayed a change with increasing bias voltage from a dual-phase structure containing cubic and hexagonal phases to a single-phase cubic structure. Investigations of the wear behaviour at various temperatures showed different controlling effects in the respective temperature ranges. The results of dry sliding tests at room temperature were independent of bias voltage and Ta-alloying, where the atmosphere, i.e. moisture and oxygen, were the most important parameters during the test. At 500 °C, bias and droplet-generated surface roughness were identified to determine the tribological behaviour. At 700 and 900 °C, wear depended on the coating’s resistance to oxidation, which was also influenced by the bias voltage. In conclusion, Ta-alloyed coatings show a significantly higher resistance to oxidation than unalloyed Ti–Al–N which could be an important reason for the improved performance in cutting operations.     

Influence of Parameters of Reactive Magnetron Sputtering on Tribomechanical Properties of Protective Nanostructured Ti–Al–N Coatings

The influence of substrate temperature and bias voltage on the structure and tribomechanical properties of the Ti–Al–N coatings obtained by reactive magnetron sputtering technique has been investigated. The structure and elemental and phase compositions have been studied by scanning electron microscopy, Rutherford backscattering, and X-Ray diffraction. The results of friction and wear experiments indicated that the lowest coefficient of friction (three times lower than 12Cr18Ni10Ti) corresponded to a coating deposited at a bias voltage of–200 V and a substrate temperature of 340°С, while the most wear-resistant coating (under a load of 700 mN and the testing time of 1080 s) was Ti–Al–N sputtered at a bias voltage of–200 V and a substrate temperature of 440°С.     

A digital nanovoltmeter

The dynamic cancellation algorithm of the temperature error of the chopped-stabilized amplifier (modulation-demodulation (MDM) amplifier) is described. The digital voltmeter, which is based on this algorithm, features a sensitivity threshold of 0.1 nV, when the time constant is 850 s, and the temperature drift of the bias voltage is 40 pV/°C in a 400-μV dynamic range.