Double-layer fuzzy adaptive NMPC coordinated control method of energy management and trajectory tracking for hybrid electric fixed wing UAVs

The energy management and trajectory tracking control are crucial to realize long-endurance autonomous flight for hybrid electric UAVs. This study aims to comprehensively consider energy management and trajectory tracking for hybrid electric fixed wing UAVs with photovoltaic panel/fuel cell/battery. A double-layer fuzzy adaptive nonlinear model predictive control method (DFNMPC) is proposed. Separated by the surplus demand power, energy management and trajectory tracking problem are decoupled into the high-layer fuzzy adaptive nonlinear model predictive controll problem (H-FNMPC) and low-layer fuzzy adaptive nonlinear model predictive controll problem (L-FNMPC). H-FNMPC solves the trajectory tracking and navigation control probelm for the greatest benefit of solar energy. L-FNMPC solves the power allocation problem of hybrid energy system for minimum equivalent hydrogen consumption. A fuzzy adaptive prediction horizon adjustment method based on UAV maneuvering degree is proposed to effectively improve proposed method adaptability to different mission profiles. Analogously, a fuzzy adaptive equivalent hydrogen consumption factor adjustment method in L-FNMPC is proposed to ensure the flexible utilization of battery. In addition, an equivalent hydrogen flow rate calculation method based on the real-time current ratio is proposed for PV/FC/Battery hybrid energy system. Numerical simulation results including a spiral trajectory tracking and a quadrilateral trajectory tracking, demonstrate that DFNMPC can simultaneously handle energy management and trajectory tracking problem for hybrid electric UAVs. Compared to hierarchical fuzzy state machine strategy, DFNMPC can save 13.3% hydrogen for the spiral trajectory tracking, and 56.9% for the quadrilateral trajectory tracking. It indicates that the energy efficiency can be improved from both levels of energy management and flight motion. The proposed method prospected for exploring high-energy-efficiency autonomous flight of hybrid electric UAVs in the future.     

Flow field simulation and pressure drop modeling by a porous medium in PEM fuel cells

For proton-exchange membrane fuel cells, the distribution of reactant flow in the stack is critical to the fuel cell's efficiency. The uneven distribution of reactant flow in the stack may cause poor current density, low performance, and material degradation. To understand and accurately predict the flow field in the proton-exchange membrane fuel cell system, the present study aims to develop a simple correlation to analyze the pressure drop in fuel cell stacks. The flow channel in each cell of a stack is treated as a porous medium, and a power-law model is used to approximate the porous medium momentum source term. For the stacks with fewer cell numbers, namely, 1, 5, and 10 cells, the parameters in the power law are established based on the experimental data. Then, a correlation is developed to simulate the flow and predict the pressure drop in the stack with higher cell numbers (ie, 20 and 40 cells). The simulations show that the pressure drop in each cell of a stack is almost invariable, and the average pressure drop decreases with increasing the number of cells. The flow uniformity in the stacks with different cell numbers is evaluated using the dimensionless pressure drop and the pressure drop ratios. It suggests that the lower the cell number, the more uniform the pressure drop. The developed model is conducive to efficiently designing the flow channel for a fuel cell stack with large cell numbers.     

Improving accuracy of automatic optical inspection with machine learning

Electronic devices require the printed circuit board(PCB)to support the whole structure,but the assembly of PCBs suffers from welding problem of the electronic components such as surface mounted devices(SMDs)resistors.The automated optical inspection(AOI)machine,widely used in industrial production,can take the image of PCBs and examine the welding issue.However,the AOI machine could commit false negative errors and dedicated technicians have to be employed to pick out those misjudged PCBs.This paper proposes a machine learning based method to improve the accuracy of AOI.In particular,we propose an adjacent pixel RGB value based method to pre-process the image from the AOI machine and build a customized deep learning model to classify the image.We present a practical scheme including two machine learning procedures to mitigate AOI errors.We conduct experiments with the real dataset from a production line for three months,the experimental results show that our method can reduce the rate of misjudgment from 0.3%–0.5%to 0.02%–0.03%,which is meaningful for thousands of PCBs each containing thousands of electronic components in practice.     

Process knowledge-based random forest regression for model predictive control on a nonlinear production process with multiple working conditions

In process industry, predictive control approaches have been widely used for nonlinear production processes. Practically, the predictor in a predictive controller is extremely important since it provides future states for the optimization problem of controllers. The conventional predictive controller with precise mathematical predictors approximating the state space of physical systems is difficult and time-consuming for nonlinear production processes, and it performs poorly over a wide range of working conditions and with significant disturbances. To address the challenges, the trend of applying artificial intelligence emerges. However, the industrial process-specific knowledge is ignored in most cases. In this study, a predictive controller with a control process knowledge-based random forest (RF) model is proposed. Specifically, working data are clustered at first to handle diverse working conditions. Then, a process knowledge-based forest predictor, namely MIW-RF model with a redesigned cascading RF structure, is proposed to incorporate control process knowledge into modeling. Thus, future states of controlled variables could be more accurately acquired for the optimizer. A simplified version of the predictive model is also developed with quick model training and updating. The proposed predictive methods are finally introduced into the controller design. According to the empirical results, the proposed methods deliver a better control performance against benchmarks, including more accurate anticipated controlled-variable responses, better set-point tracking and disturbance rejection capability.     

大豆蛋白纳米纤维对铁纳米颗粒的稳态化作用

为明确大豆蛋白纳米纤维的结构形成和扩宽铁强化剂的食品工业应用，以大豆分离蛋白（soy protein isolate，SPI）为原料，通过5 h的酸热处理制备纳米纤维（soy protein isolate fibrils，Fib SPI），系统研究纤维形成前后蛋白结构的变化，并进一步制备铁纳米颗粒（iron nanoparticles，Fe NPs），探究Fib SPI对铁的稳态化作用。研究结果表明：在酸热处理过程中，SPI产生大量的β-折叠结构，其与硫磺素T结合，显示出增强的荧光强度；此外，7S组分先发生降解，利于纤维成核形成，随后11S逐渐被水解，促进纤维生长；同时水解产生大量的小肽组分，提高了产物的还原力。研究进一步利用Fib SPI递送铁纳米颗粒（Fe NPs），发现与原始SPI相比，铁纳米颗粒可在Fib SPI原位形成胶体稳定的铁-大豆蛋白纳米纤维复合物（Fe FibSPI），并以Fe（II）形式存在，其对乳液体系色泽及稳定性的影响较硫酸亚铁或氯化铁小。该研究可为构建新型植物基铁强化剂递送体系提供理论和方法指导。     

基于VMD和BSA-KELM的高陡边坡位移预测模型研究

边坡位移的时间序列曲线存在复杂的非线性特性,传统的预测模型精度不足以满足预测要求。为此提出了基于变分模态分解的鸟群优化-核极限学习机的预测模型,并用于河北省某水泥厂的边坡位移预测。该方法首先采用VMD把边坡位移序列分解为一系列的有限带宽的子序列,再对各子序列分别采用相空间重构并用核极限学习机预测,采用鸟群算法优化相空间重构的嵌入维度和KELM中惩罚系数和核参数三个数值,以取得最优预测模型。最后将各个子序列预测值叠加,得到边坡位移的最终预测值。结果表明：和KELM、BSA-KELM、EEMD-BSA-KELM模型相比,基于VMD的BSA-KELM预测精度更高,为边坡位移的预测提供一种有效的方法。     

Thermal properties of hierarchical YSZ and LZO ceramic microspheres with multi-scaled voids investigated by using theoretical,experimental and simulation methods

Heat transfer within ceramic feedstock powders is still unclear, which impedes optimization of the thermal and mechanical properties of the thermal sprayed coatings. The microspheres (yttria-stabilized zirconia YSZ and lanthanum zirconate LZO) were prepared via the electro-spraying assisted phase inversion method (ESP). The thermal properties of the two ESP microspheres and a commercial hollow spherical powder (HOSP) were investigated by using theoretical, experimental, and simulation methods. Thermal conductivity of the single microsphere was estimated via a novel nest model that was derived from the Maxwell-Eucken 1 and the EMT model. Thermal conductivity of a single YSZ/LZO-ESP microsphere prepared at 1100–1200 °C was within 0.36–0.75 W/m K, which was ～ 20 % lower than that of a single YSZ-HOSP microsphere with a similar porosity. Heat flux simulation showed that high tortuosity around the multi-scaled voids of the ESP microsphere led to a more efficient decrease in thermal conductivity compared with total porosity.     

基于平均飞行航迹的CCO减噪效用研究

近年来，通过优化飞行程序降低机场飞机噪声影响成为机场环境保护的重要研究方向。文章首先建立了基于飞机“噪声-功率-距离”数据的噪声计算模型，介绍了平均飞行航迹以及连续爬升运行(Continuous Climb Opera-tion, CCO)离场程序的相关理论，最后以大型国际机场为实例，使用飞机平均飞行航迹进行噪声预测，运用综合噪声模型计算出噪声影响面积并绘制噪声影响等值线图，比较了CCO离场相对常规的标准仪表离场(Standard Instru-ment Departure, SID)的降噪效果。结果表明，CCO离场程序可有效降低机场噪声影响，在高噪声级影响区域的降噪效果更佳。     

两种混合单元混合机理的数值模拟

为了研究SV型和SX型混合单元在全静态乳化器的混合机理,使用群体平衡模型对乳化器内部流场进行模拟分析。结果表明,SV型与SX型混合单元的混合机理是油相与水相材料通过混合单元时其空间流动方向改变,在流场内形成大量的漩涡,从而达到乳化的目的;通过比较流体在经过SV型和SX型混合单元后的水相索特平均直径(d_(32)),得出SV型混合单元的乳化效果更强,但同时消耗更多的能量,压降更大。