A 6-DOF humanoid wall-climbing robot with flexible adsorption feet based on negative pressure suction

A wide range of dangerous and special tasks have witnessed the applications of wall-climbing robots, but they still cannot adapt well torough or sloping walls. This paper proposes a 6-DOF (degree of freedom) humanoid wall-climbing robot (HWCR) based on the principle of negative pressure suction. HWCR has the advantages of flexible adsorption feet, strong adaptability, strong anti-subversion performance, and high friction to the wall. We deduce mechanics formulas and carry out a parametric design of the foot structure so that it can meet the requirement of robot wall climbing. We use Fluent to analyze the flow field of the adsorption foot and determine the motor speed that can provide a reliable adsorption force. Using the D-H matrix to plan gait, we also design a compound cycloid-based foot trajectory to reduce the impact between the HWCR and the wall. Experiments on the uneven wall and sloping wall show that the vehicle can walk with an ideal gait, and the resistance value of the servo on each joint is much lower than the critical value, which ensures the smooth movement of the HWCR.     

SOINN-Based Abnormal Trajectory Detection for Efficient Video Condensation

With the evolution of video surveillance systems, the requirement of video storage grows rapidly; in addition, safe guards and forensic officers spend a great deal of time observing surveillance videos to find abnormal events. As most of the scene in the surveillance video are redundant and contains no information needs attention, we propose a video condensation method to summarize the abnormal events in the video by rearranging the moving trajectory and sort them by the degree of anomaly. Our goal is to improve the condensation rate to reduce more storage size, and increase the accuracy in abnormal detection. As the trajectory feature is the key to both goals, in this paper, a new method for feature extraction of moving object trajectory is proposed, and we use the SOINN (Self-Organizing Incremental Neural Network) method to accomplish a high accuracy abnormal detection. In the results, our method is able to shirk the video size to 10% storage size of the original video, and achieves 95% accuracy of abnormal event detection, which shows our method is useful and applicable to the surveillance industry.     

Multi-objective optimal allocation strategy for the energy internet in Huangpu District,Guangzhou, China

To improve the overall efficiency of the energy system, the basic structure for the energy internet of coordination and optimization of “generation-grid-load-storage” of Huangpu District, Guangzhou, China is designed, while the arrangement for the output of centralized and distributed energy module and energy storage are proposed. Taking economic benefit maximization, environmental benefit maximization and energy efficiency maximization as sub-objectives, the mathematical model of multi-objective optimal allocation and operation strategy of the energy internet is established considering supply-demand balance constraints, equipment characteristic constraints, operation mode constraints, and energy conditions constraints. The calculation results show that without considering the outsourced electricity, the balanced strategy, the economic development strategy, the environmental protection strategy, and the energy efficiency strategy are obtained by calculation, which are all superior to the traditional energy supply strategy. Moreover, considering the outsourced electricity, the proportion of outsourced electricity to total electricity is 19.8%, which is the system optimization of the energy internet under certain power demand. Compared with other strategies without outsourced electricity, the outsourced electricity strategy can have a certain emission reduction effect, but at the same time reduce the economic benefit. Furthermore, the huge difference in demand for thermal and cooling load between industrial and commercial areas results in the installed capacity of gas distributed energy stations in industrial areas being nearly twice as large as that in commercial areas. The distributed photovoltaic power generation is allocated according to the proportion of the installed roof areas of photovoltaic power generation system in residential, industrial, and commercial areas.     

基于深度学习的传感器优化布置方法

针对传感器优化布置(optimal sensor placement,简称OSP)问题,提出了一种新的使用深度神经网络的解决方案,并以简化的桥梁形状的桁架结构中的振动测试传感器优化为例进行了验证。首先,选择一种传统的传感器优化布置方法,对自动化生成的大量不同的桁架结构分别进行传感器优化布置计算,将所得优化布置结果在进行数据预处理后构建出深度学习方法所需要的训练集与验证集;其次,使用Python语言和深度学习框架TensorFlow设计实现与本研究问题适配的深度神经网络模型并训练;然后,随机生成了新的桁架结构参数;最后,将深度神经网络输出的传感器布置结果和传统方法的计算结果进行了比较,验证了本研究方法的有效性以及在速度上、可移植性与可扩展性方面的性能优势。     

枞树型核电轮槽精铣刀制造技术优化设计

以枞树型叶根精铣刀为例,阐述了可换硬质合金型线刀体和刀片精铣刀的优化设计、制造工艺、建模及优化设计,进行切削过程加工策略优化和加工过程模拟,设计制造了锥柄专用工装及检测环规、磨削刀片后角专用工装,采用NUMROTO软件编制刀片加工工步。该项目研制的可换刀片精铣刀型线精度完全能达到进口铣刀水平,能满足大部分轮槽与叶根加工对型线精度的要求,可以大幅度节约刀具费用。     

基于近似测地线的分层次自动铺带轨迹规划方法

自动铺带技术是一种增量制造技术，适用于制造翼面、壁板等大尺寸、小曲率复合材料构件，具有加工成本低、效率高等特点。轨迹规划是自动铺带技术的关键环节，它直接关系到复合材料构件的成型精度与质量。现有的自动铺带轨迹规划方法难以便捷兼顾调控铺放间隙与防止铺放褶皱，保证铺带产品的最终质量。有鉴于此，提出了一种基于近似测地线的分层次铺放轨迹规划方法。主要目标是合理利用预浸带允许变形能力，在防止铺放褶皱产生的前提下，保证相邻带料不产生覆盖，且间隙不超过2.5 mm。首先，研究了铺放间隙的产生与演化规律，给出了几种典型曲面上初始铺放间隙的优化方法。接下来，提出了近似测地线的数学模型与求解格式。然后，基于铺放褶皱的产生机制与铺放间隙的演化规律，确定了近似测地线的测地曲率大小与符号。最后，在双曲面模具以及自由曲面模具上验证了所提出方法的有效性。     

干湿联合冷却塔冷却节水分析软件开发及应用

为实现对干湿联合冷却塔冷却特性、节水特性的综合分析，编制干湿联合冷却塔校核计算流程，基于Visual Studio开发平台，本文开发了干湿联合冷却塔冷却节水特性分析优化软件。通过对某在运干湿联合冷却塔进行冷却节水特性计算，验证了所开发软件计算结果的准确性；在消雾节水计算结果基础上，对标干湿联合消雾冷却塔验收测试规程，生成其成雾频率曲线、耗水量曲线、塔雾指数、冷却特性曲线，综合分析其消雾特性、节水特性及冷却特性，从而确定了干湿联合冷却塔的最佳运行工况。同时运用所开发软件对比分析了某冷却塔设计干段散热面积对其消雾特性、节水特性及冷却特性的耦合影响，并分析了百叶窗开度对干湿联合冷却塔性能的影响，为干湿联合冷却塔的运行优化和设计优化提供了工具性软件支持。     

论三峡水库“蓄清排浑”运用方式及其优化

泥沙问题是三峡工程的关键技术问题之一,在三峡工程论证和初步设计阶段提出水库采取"蓄清排浑"的运用方式,可以解决泥沙问题。2003年水库蓄水运用以来的实践表明,水库基本遵循了"蓄清排浑"调度运用原则,并根据上游来水来沙减少等新情况对水库运用方式进行了优化调整。本文系统分析了三峡水库"蓄清排浑"运行方式及其优化调整的利弊,包括提前5年实施175 m试验性蓄水、汛期水位动态变化、汛末提前蓄水等对水库淤积和坝下游河道演变的影响,提出了进一步优化水库调度方式,形成"蓄清排浑"运用的新模式,保持水库长期使用的建议,试图为三峡水库科学高效安全运用提供科技支撑。     

建筑材料供应中断对承包商成本的冲击研究——基于鲁棒模型分析

为了揭示供应商与承包商为主体的二级供应链中，两者间耦合作用条件下发生的供应中断对供应商最优决策及对承包商成本的影响机制，且剖析供应中断对承包商成本的冲击效应，以供应商存在高退出风险为研究对象，在仅知产量均值与方差条件下，运用Scarf提出的“极大极小值”鲁棒决策方法，通过设定不同决策情况下的供应链系统反馈，分析了不同退出标准情况下二级供应链的博弈过程。基于理论支撑下的鲁棒模型数值仿真结果得出：在退出标准α为100时，中间变量在290~350范围内，供应商存在最优决策，但承包商最优决策无法确定；与α为0时相比，在订货量达到一定状态，即α为140时，承包商成本增加了1.5%，表明承包商成本与供应商退出标准呈正相关，退出标准越高，对承包商成本造成的冲击效应越大。此研究结论为供应链初期不稳定且信息不完整条件下发生的供应中断影响机理提供了新的视角。