To solve the position control problem of the induction motor with parameter perturbations, load disturbances, and modeling errors, a predefined-time position tracking optimization control method with prescribed performance is proposed. In practice, the rotor flux linkage of the induction motor can't be measured, and a predefined-time sliding mode observer (PTSMO) is applied to accurately estimate it. Additionally, predefined-time disturbance observers (PTDOs) are employed to identify the uncertainties in the motor system. The position and flux linkage controllers are then designed by integrating the predefined-time control approach with the prescribed performance function method, realizing accurate tracking control of the induction motor within a predefined time. Next, the adaptive genetic algorithm (AGA) and the improved particle swarm optimization (IPSO) technique are combined to optimize the designed controllers, enhancing the convergence rate and steady-state accuracy of the induction motor. Finally, comparative analyses through simulations and dSPACE simulated experiments validate the efficacy of the proposed control method, highlighting its applicability in practical motor systems.
物联网设备之间通过不同的通信协议进行数据传输与信息交互.针对设备异构通信协议相互理解问题,深入研究了多工业协议解析技术,研究并实现了一种可支持协议扩展的语义解析网关.网关定义了多种工业协议解析脚本、开发了词法分析和语法分析程序、生成对应的协议解析指令集和实现了解析脚本执行程序和解析数据输出,建立了整体的语义解析系统.网关通过定义不同协议的解析脚本对不同协议的数据报文进行语义解析,析出的语义数据可以直接被设备或者云端应用,从而实现了更加灵活的数据传输和信息交互.网关通过扩展解析脚本来支持对协议扩展,不需要针对每种协议单独设计解析程序,极大地减少了协议解析程序开发的难度和工作量,更符合当前及未来物联网发展的趋势.通过对网关性能进行可靠性和实时性测试实验,验证了网关对工业协议解析转换稳定可靠及时.
In the context of sharing economy, establishing a stable alliance is crucial in the logistics service industry. Consequently, the challenge of ensuring fair profit allocation arises within the coalition of logistics enterprises. However, due to the presence of incomplete information in the coalition, some traditional point-valued solutions of cooperative games, such as the Shapley value, may be inadequate. These solutions are more proper for cooperative games where accurate estimate of both the general profit of the alliance and the participation level are feasible. In this study, we address the issue of profit allocation for logistics enterprise coalitions with incomplete information and propose a relevant profit allocation model. To demonstrate the applicability of the proposed model, a case study is provided. The results show that the fuzzy Shapley value significantly enhances the multi-party cooperation and can serves as an effective tool for the fair and equitable allocation of profits within logistics enterprise coalitions.
Aiming at the weak generalization ability of traditional rolling bearing fault diagnosis methods, a rolling bearing diagnosis method based on dynamic convolutional capsule networks (DC-CapNets) was proposed. First, one-dimensional vibration signals are preprocessed and divided into a training set and a test set. Then the fast Fourier transform (FFT) is used to convert the training set and the test set. In this model, two dynamic convolution layers and pooling layers are used to extract the features of the input frequency signals, and the feature information is transmitted to the capsule network to obtain the diagnosis results. The CWRU (Western Reserve University) data set validation shows that the proposed method still has high fault diagnosis accuracy under noise and variable load conditions, and has good anti-noise and generalization performance, which is superior to the convolutional neural network (CNN) and deep neural network (DNN).
Addressing the challenge of rehearsing large-scale equipment assembly, particularly for oversized components like wind turbine towers, hubs, nacelles, and blades, which often face quality issues such as substandard workmanship and wide tolerances during on-site assembly, a system has been developed for simulating wind turbine assembly. This system enables digital wind turbine assembly by creating virtual and production process models and employing intelligent database analysis. It resolves the problem of the absence of pre-production for large-scale equipment, meets the batch production needs of the wind turbine manufacturing industry, enhances the safety of wind turbine operations, improves operator assembly skills, and boosts production efficiency. This platform has already been implemented in the wind turbine manufacturing industry, yielding significant economic benefits.
In order to make the ocean energy generator system realize the maximum power under various conditions, this paper proposes a control strategy based on linear active disturbance rejection control algorithm. The output power is mainly affected by the spin-speed of the generator in a nonlinear manner, and this phenomenon is originated from the tip speed ratio state of the turbine blades. This tip speed ratio associated maximum power can be achieved through a coupled control system that adjusts the spin-speed of generator elaborately. The state equations of ocean power generation system are transformed into a standard form to fit the linear active disturbance rejection control. Finally, the simulation model of the system is established by using Matlab/Simulink platform, with a series of input signals for different working conditions. The results verify that the maximum power under control can be effectively attained by a quick tracking of the target speed.
"智能工厂"是工业自动化的必经之路,不同企业的生产流程、任务协作具有其自身的特性.在此种环境下,结合企业自身条件、资源环境、生产特性对工厂的人员、软硬件进行统筹调度,是实现智能化、自动化亟待解决的制造痛点.基于表面贴装(SMT)机器加工车间,在深入研究车间的人员管理、生产流程、物流运转的前提下,基于面向服务(SOA)架构理念的基础上开发智能工厂连携系统,利用库位管理算法结合任务调度算法制定智能移障算法逻辑,实现移动障碍搬运任务目标并且将障碍物移动进入合理的任务点位.充分融合二维码扫描进行物料校验的物联网(IOT)技术、Windows通讯开发平台(WCF)与安卓网络框架之OKhttp进行服务通信的计算机通信技术、实时改变任务步骤与硬件动作地点路径的数据分析处理技术,达成生产计划自动执行、无人化物流、生产数据自动传输、智能移障功能,形成智能加工供应链,实现SMT加工车间的信息化和智能化应用落地,为企业智能制造赋能.
在室内环境中,移动机器人精确的位姿估计是实现机器人控制的先决条件.为了解决在稀土冶炼电解车间移动机器人使用单一传感器存在定位误差大、精度低、稳定性差的问题,提出了一种将惯性测量单元(IMU)、轮式里程计和激光雷达等多种传感器数据融合的定位方法.该方法将轮式里程计和IMU的位姿信息通过无迹卡尔曼滤波算法融合,然后将融合后的数据与激光雷达信息通过粒子滤波算法融合,得到更精确的位姿数据.经过实验论证表明,该方法可以将定位误差稳定到2cm左右,显著提高了移动机器人的定位精度和鲁棒性.
为解决烟箱码垛后表面缺陷未被检测问题,设计一种烟箱表面图像检测模型.通过提取烟箱表面缺陷特征,形成训练集和与烟箱分类器,并基于机器视觉阈值,引入阈值差比对检测算法,代替传统的人工缺陷检测方法.通过图像预处理、噪声处理、图像特征选择和提取等方法,使阈值和烟箱正向图像进行比对,智能判断烟箱表面是否存在缺陷.改进后,机器人在码垛过程中,能够准确地识别出烟箱表面缺陷并能及时将表面有缺陷的烟箱剔除,准确率达到97%.实验结果表明,烟箱表面缺陷检测模型可满足高质量生产需求,具备优良泛化性能.
借助现代信息技术手段,解决传统人工在矿山执法监察过程中存在的劳动强度大、执法效率低、信息化程度不高等问题,用以全面提高矿山执法监察的能力与水平,是当前新形势下的必然选择.以云南省矿山为研究对象,综合采用无人机航测等技术获取的矿山实景三维模型、真正射影像(TDOM)、数字高程模型(DEM)和数字线化地图(DLG),研发的一套"空天地"一体化矿山动态执法监察系统,更新并夯实矿山基础数据库内容,并使用系统功能模块,以实现矿产资源监察为目标,对各类违法开采事件起到高效、科学、准确的监管,更好的辅助矿区管理与人工执法,维护国家利益,保障人民群众生命财产安全.
为了推进电网资产实物"ID"规模化应用,考虑到当前电网资产实物"ID"标签制作现状所面临的挑战,提出了基于国密算法的电网资产实物"ID"标签属地化制作模式.在应对这种模式所带来的信息安全风险时,设计了以Tracking No.为分散因子基于国密SM7算法生成标签EPC区Access口令,基于国密SM1算法生成密文Tracking No.和密文Access口令进行主生产线和属地生产线之间的数据远程传输.最后,本文完成了电网资产实物"ID"标签属地化制作模式的设备部署,结合国密算法的安全机制,可实现电网资产实物"ID"标签属地化制作.
自动码垛机器人系统在物流自动化领域得到了广泛的应用,然而,在民航行李处理系统的应用并不多见,主要是旅客行李尺寸的无规则性、行李材质的多样性、行李发生的随机性等原因导致其应用困难.面向民航行李处理系统托运行李装车场景,为提高工人装车效率,降低工人劳动强度,搭建工业机器人视觉引导的行李码垛系统.基于3D相机、工业机器人、输送设备、上位控制软件,提出系统构成、设备选型、工艺流程与上位控制软件设计.通过坐标系融合获得3D行李点云图像在工业机器人体系中的坐标,并对3D行李图像进行降采样与去噪处理.工业机器人根据3D相机获取的行李位姿信息进行码放,对偏离计划位置的行李进行矫正.实验结果表明,3D视觉引导工业机器人行李码垛系统的机器人漏抓率为0,位置码放一次准确率为86.27%,矫正后位置准确率为100%,验证了系统的实用性.
针对高速卫生纸机抄前池液位和匀整磨出口流量存在耦合性、非线性、时变性的问题,基于进出抄前池纸浆绝干量动态平衡原理,利用多义线性方程求取抄前池液位与匀整磨出口流量的修正系数,以稳定匀整磨出口流量的方式,实现抄前池液位恒定.设计双模糊PID流量控制算法,依据流量偏差和变化率不同选择合适的模糊控制算法,对PID参数进行实时整定,以提升控制系统的响应速度、稳定性及精确性.以SIE-MENS CPU414-3PN/DP为核心控制器,PCS7集成软件为控制系统开发平台,完成系统硬件组态、软件开发、HMI界面设计,并将该系统成功应用于保定市某高速卫生纸机DCS控制系统.现场运行和测试:抄前池液位恒定误差为1.36%、匀整磨出口流量误差为0.18%,表明该控制方案具有较好的可行性和适应性.
针对某司锻轧线车轮轧机生产数据本地保存的需求,基于TIA portal WinCC,开发脚本,将轧机生产数据实时存储到本地计算机的Excel报表中,如车轮轧制时间、车轮外径、车轮温度以及压力、转速等数据,并且数据报表在本地计算机保存120天.
针对医药电商物流中心和智能药房中高速自动发药机的发药准确性问题,提出了复杂光场下基于视觉的药盒快速计数算法.该算法包括图像预处理、目标提取、目标计数三个步骤.首先,在图像预处理中采用色彩空间转换、中值滤波、Canny算子和大津算法完成图像二值化.之后,在目标提取中融合轮廓检测和连通域提取两个算法获取候选目标.最后,在目标计数中通过轮廓过滤和去重处理得到最终计数结果.实验结果表明,该方法具有较好的计数精度及识别速度,为药盒自动化计数研究提供了新的方法.
提出一种基于轨迹灵巧度指标的机械臂尺寸优化方法,能够提高机械臂在指定轨迹上的运动灵活性.在运动学分析、雅可比矩阵求解和笛卡尔空间轨迹规划的基础上,以轨迹条件数与轨迹条件数波动系数加权值为灵巧度指标,采用遗传算法对机械臂尺寸进行了优化.优化结果显示:优化后机械臂轨迹灵巧度提高了17.5%,同时机械臂全域灵巧度提高了11.0%.
为了提供车削碳纤维复合材料的切削力及表面质量的数据,提出了超声振动车削的方式提高表面质量,对超声振动切削的运动以及切削力模型进行分析后,自制超声振动状态下的切削力检测刀具,研究不同的进给量、背吃刀量、切削速度对振动切削力的影响,以及不同参数切削后的表面粗糙度.在对碳纤维复合材料的破坏机理进行研究后分析切削力及表面质量得到:随着进给量以及切削深度的增加振动切削力呈现增大趋势,切削速度的变化对切削力影响较小.同时现切削参数在切削深度0.5mm,进给量0.11mm/r以下时,表面质量较好,可以作为发动机检测切削参数.
随着电网数据采集系统的不断完善,对电网二次设备内部参数进行量测已成为可能,弥补了现有二次设备健康状态评估方法实用性差的缺陷.在相关理论基础上,对二次设备健康状态评估方法进行优化、改进.首先,在实时监测的动态数据和运维积累的静态数据基础上,利用区间型主成分分析法(PCA)综合构建评估对象的评价指标体系;其次,以乘法集成法组合由区间型序关系分析法和熵权法分别确定指标的主、客观权值;同时,基于云模型构造的状态空间计算基础层指标隶属度;然后,引入模糊合成算子实现基础层指标隶属度向目标层指标隶属度的转换,并根据综合决策方法判定设备的最终状态;最后,以某智能变电站的继电保护装置为例,对其健康状态进行评判,并加以量化分析.
针对传统矩阵算法在故障报警信息畸变时容错性差,而智能优化算法在处理大型配网时效率低且易陷入局部最优解的问题,提出一种矩阵算法与混沌二进制粒子群算法(CBPSO)协同的配电网故障定位方法.首先,基于配电网实际结构建立区段与节点的因果关联矩阵与判据,当满足因果判据时运用矩阵算法即可对故障区段实现准确快速定位;其次,当告警信息存在畸变而导致因果判据不满足时,根据矩阵算法的结果可确定故障可疑区段集合,基于该集合建立定位优化模型从而大大缩减了后续待求解变量的维数;最后,采用全局搜索能力更强的混沌二进制粒子群算法对模型进行求解,避免了二进制粒子群(BPSO)易早熟收敛的问题.仿真结果验证了所提方法具有准确、高效的优点,适用于大型配电网的故障区段定位.
分布式发电系统接入配电网容易造成节点电压的波动,影响用电设备的工作.随着电池技术的成熟,通过电池储能系统(Battery energy storage system,BESS)控制电池充放电实现配电网电压调节的方法得到了广泛的应用.由于电池充放电次数有限,提出一种节点电压排序调节算法以减少电池的充放电次数,提高电池的使用寿命,并根据节点电压值设定储能系统充放电阈值,使未越限节点辅助参与电压调节.并检测电池的电荷状态(State of Charge,SOC),根据SOC选取合适的充放电组合,避免出现过充、过放问题.利用IEEE 33节点配电网模型对提出的调节方法进行了仿真验证,结果表明所提出的调节方法能够实现电压的调节,并降低电池的充放电次数.
