贴几个 MapleSim 和 Maple 在机器人技术领域 应用 资源转自 SIMWE 网络会议录像: ) N* ` ?$ `6 Y8 s! ^ Robotics Applications with MapleSim and Maple- c2 `0 O2 ^3 n7 L! a7 _ Developments in advanced model formulation techniques and efficient model simulation algorithms, together with ever-increasing computational power, have enabled the proliferation of advanced robotics research and applications in the areas of humanoid robots, autonomous mobile robots and space robotic systems. MapleSim is one of the most advanced tools in modeling and simulation for modern robotic systems. This webinar will illustrate how MapleSim users can leverage its unique symbolic advantage to easily obtain 3D animation of the system dynamic simulation, access the full set of symbolic system equations for design and analysis, such as forward and inverse kinematics and dynamics analysis, and generate efficient real-time capable models for Hardware-In-the-Loop (HIL) applications. We will also demonstrate how conventional modeling techniques, such as D-H parameters and homogeneous transformations, can be integrated within the MapleSim physical modeling environment. 2 k# i5 A/ F; Y+ N9 x4 R' o 用户案例节选: " x3 x ~$ w% E; N3 s6 z9 s/ m0 g1 F 用户案例:类人机器 人 学 习 更快行走 2 z5 U' q k$ E$ H; x( Z 曼彻斯特 CICADA 中心,由 Martin Brown 和 Gustavo Medrano-Cerda 领导的项目组,面临的一个挑战是,如何快速和有效地可视化实验,避免影响 工程 进度,并确保实验的有效性和相关性。 “MapleSim 的可视化 功能 ,无需编写自 定义 程序 ,对我们非常有价值, ”Houman Dallali 博士说, “ 更重要的是,我们可以直接生成 C++ 代码到硬件接口,加速控制器实现 / 调试的进程。 ” * e" I e" Q! n/ j - q6 U$ z! w* F 3 D" q* G# m# M" w t' c+ L 用户案例: Automation, Robotics, and Mechatronics Lab at SUNY Buffalo 在高级研究项目中使用 MapleSim 和 Maple * B0 `/ S! H3 M! HK ARM 实验室的一个研究项目中包含对 6 自由度并联机构( 6-Prismatic-Universal-Spherical ) 运动 学和动力学 仿真 的研究。该机构包含一个移动平台、一个固定基座、数个相互连接的支脚。在研究中, Dr. Venkat Krovi 和课题组同事分析了一个 6 自由度并联机构( 6-Prismatic-Universal-Spherical ),他们使用 MapleSim 和 Maple 自动生成系统的控制方程、在 Maple 中完成运动学分析。 - ]8 eD, a# k9 j 用户案例:日本早稻田大学类人机器人研究 http://www.maplesoft.com/company/publications/articles/view.aspx?SID=370 : \, N x7 {7 B: J$ d
Modeling a Double Pipe Heat Exchanger with MapleSim Posted: December 29 2010 by Samir Khan 259 Products: Maple MapleSim 6 0 A prospective customer recently asked if we had a MapleSim model of a double pipe heat exchanger. Heat exchangers are a critical unit operation in the process industries, and accurate models are needed for process control studies. I couldn't find an appropriate model so I decided to derive the dynamic equations, and implement them using MapleSim's custom component interface. I'll outline my modeling strategy in this blog post. Typically, double pipe heat exchangers are modeled using a continuum approach, in which the temperature variation across both streams is described by PDEs, or a discretized approach, in which the temperature variation is described by ODEs. Given that MapleSim solves ODEs (and not PDEs), I chose the discretized approach. 1 Deriving the System Equations 1.1 Introduction The heat exchanger was divided into N control volumes. A heat balance on a typical control volume resulted in three differential equations – one each for the tube- and shell-side liquid, and one to model the heat capacity of the tube wall. Axial heat flow along the tube wall (heat flow into and out of tube wall sections due to temperature differences in adjacent tube wall sections) was also modeled with Fourier's Law of conduction. For the simple model outlined here, I've assumed that heat exchanger was insulated, so no convective heat losses from the surface are considered. 1.2 Energy Balance on the Tube and Shell-Side Streams For a single control volume, a heat balance on the tube-side stream gives However Ttin,iand Ttout,i(the inlet and outlet tube-side temperatures in each control volume) are not state variables. They are approximated by taking the average of adjacent temperatures: Hence the heat balance on the tube-side becomes A similar heat balance on the shell-side fluid gives 1.3 Energy Balance on the Tube Wall The tube wall acts as a heat capacitor, and can have a significant effect on the transfer of energy from one stream to another. It is assumed that the tube wall has a homogenous temperature in each control volume, with heat transferred to and from the tube- and shell-side liquids, and via conduction from adjacent tube-wall sections. A heat balance gives: In reality, the inner and outer surface would have different temperatures. This could be modeled by dividing the tube wall into several layers, performing a heat balance on each with Fourier's Law governing heat flux between layers. This simplified model ignores this effect. 1.4 Heat Transfer Coefficients The heat transfer coefficients hwt and hws were predicted by the Dittus-Boelter correlation . 2 Implementation of System Equations in a Custom Component The entire set of differential equations for all N control volumes were generated in a MapleSim custom component with a simple application of the ?seq command (this approach meant I could explore how increasing the number of control volumes affected the results simply by changing the value of N and regenerating the custom component). Figure 3 give a small subset of equations implemented in the custom component (see the attached MapleSim model for the full set), while Figure 4 outlines the steps involved in creating a custom component. It's important to note that no causality has been specified, so the block can be used in any configuration. 3 Complete Heat Exchanger Model The heat exchanger as outlined above can be downloaded here . A more sophisticated version of this model can be found here . This version accounts for the temperature variation of the tube-side liquid viscosity (which can have a significant effect on the tube-side heat transfer coefficient, as predicted by the Dittus-Boelter correlation), and also implements a temperature control loop. Another version (that I can't share) models the temperature variation of the tube wall to a greater fidelity, and convective losses from the shell surface. 4 Notation Dit Inside diameter of tube m Dot Outside diameter of tube m Dis Inside diameter of shell m t Density of tube-side fluid kg m-3 s Density of shell-side fluid kg m-3 w Density of tube-wall material kg m-3 Cpt Specific heat capacity of tube-side liquid J kg-1K-1 Cps Specific heat capacity of shell-side liquid J kg-1K-1 Cpw Specific heat capacity of tube-wall material J kg-1K-1 hsw Heat transfer coefficient of shell-side fluid and tube wall W m-2K-1 htw Heat transfer coefficient of tube-side fluid and tube wall W m-2K-1 L Length of heat exchanger M N Number of control volumes Tti Temperature of tube-side liquid in control volume i K Tsi Temperature of shell-side liquid in control volume i K Twi Temperature of tube wall in control volume i K Qt Flowrate of tube-side liquid m3s-1 Qs Flowrate of tube-side liquid m3s-1 maple maplesim applications Add Comment Branch Contact Author Flag This 2403 views You must be log into your MaplePrimes account in order to post a comment. If you don't have an account, you can create an account here . User Name Or E-mail: Password: Remember Me: Automatically sign in on future visits Forgot Your Password? Create an Account
NCTUns 5.0 安装成功 毕业论文需要做 IEEE 802.11p 的仿真,选择了 NS-2 和 NCTUns 作为工具,在安装 NCTUns 时出现和解决了一些问题,终于于 2010 年 3 月 20 日正式装好 NCTUns 软件,安装过程颇为心酸,特撰此文留念。 3 月 15 日,初步打算使用 Windows XP+Ubuntu 9.10+ VMware 6.02+NCTUns 6.0 安装,然而经过 3 天的折腾后,发现遇到的问题越来越多,到 NCTUns 官方论坛上荡了一圈,发现遇到同样问题的同志们不在少数,毅然于 2010 年 3 月 18 日转战 Fedora 10 ,大家可以从 Fedora 的官方网站上下载 iso 镜像文件,这里提醒大家,根据自己需要, 一定要选对 i386 或 x86 , 否则可能导致安装失败! 最终选择了 Windows XP+Fedora 10+ VMware 6.02+NCTUns 5.0 首先用 VMware6.02 安装好 Fedora 10 ,具体步骤大家可参看网上相关教程或给我留言。 安装必备软件包: 如果你的 Fedora 能联网,就使用类似下面的命令更新包 # yum update # yum install gcc gcc-c++ make mkinitrd 如果不能联网,请从前述下载的 iso 镜像文件中 (Fedora-10-i386-DVD.iso-3.41G) ,拷取 package 中相应的 gcc 包和 gcc-c++ 包,需要说明的是,如果你想安装 NCTUns 6.0 ,最好使用 Fedora 12 ,否则需要将 gcc-c++ 更新至更高的版本。 拷出相应包后,安装 rpm 包及相关的依赖包: #rpm ivh packagename.rpm 本人一共大概安装了 readline-devel 、 ncurses-devel 、 gcc 、 gcc-c++ 、 glibc-devel 、 glibc-headers 、 kernel-headers 、 libstdc 、 ncurses-devel 、 readline-devel 、 rpm-libs 等包 安装好常用包后,我们就可以着手安装 NCTUns 了 下载 NCTUns5.0-allinone-linux-2.6.27.7-f10.20090709.tar.gz( 我好像改过文件名了 ~) ,解压缩并运行安装程序: # tar zxvf NCTUns5.0-allinone-linux-2.6.27.7-f10.20090709.tar.gz # cd NCTUns-5.0/ # sh install.sh 大概 5-6 分钟后,安装完毕,我们重新启动 # reboot 重启, 选择 NCTUns 内核 ,运行 NCTUns 会出现下面问题: failed to load the logo 以及 cannot find the file containing the module list 的提示 大家注意:这里我们 必须以 root 身份运行 ! 切换到 root 身份,重新运行 NCTUns ,熟悉的 GUI 出现了 如果重启后运行 NCTUns 出现下面问题: $ /usr/local/nctuns/bin/nctunsclient /usr/local/nctuns/bin/nctunsclient.bin: error while loading shared libraries: libqt-mt.so.3: cannot open shared object file: No such file or directory 则需要修改 .bashrc 配置文件: $ vi .bashrc export NCTUNSHOME=/usr/local/nctuns export NCTUNS_BIN=/usr/local/nctuns/bin export NCTUNS_TOOLS=/usr/local/nctuns/tools export LD_LIBRARY_PATH=/usr/local/nctuns/lib 提醒大家:以 root 用户打开 run simulation 时, GUI 所需密码默认是 nctuns 。请注意区别! ( 附安装后常见问题: 1 、 How to solve login failed. please check your username and password problem It is because the account/password doesn't match with the user information in the dispatcher setting. To solve the problem, run the nctunsclient and then goto G_Setting - Dispatcher. Under user information, give an ordinary username and its password (the account can't be root). Of course, the ordinary user account should be previously created in your linux machine. 2 、 How to solve No Idle Server error The no idle server shows up may be the reason that the program terminates unexpectedly on the previous simulation. You can use ps aux | grep nctuns to see what programs are still running and then use killall to kill that process. Example: killall stg 3 、出现了 Connect to Dispatcher at 127.0.0.1:9800 failed , 执行 Dispatcher, ,还要另外开启 coordinator 才能 run simulation ,注意,要开三个终端,其中一个开 Dispatcher ,一个开 coordinator ,一个执行 Nctuns 4 、以 root 身份登录 fedora 10 编辑 /etc/pam.d/gdm ,注释或删除: # auth required pam_succeed_if.so user != root quiet 有时可能还要编辑 /etc/pam.d/gdm~ 注释或删除: # auth required pam_succeed_if.so user != root quiet 保存文件后,登出 GUI 后,即可用 root 登录 GUI 了。 5 、安装文件中 readme 关于环境的设置 Before running all of the NCTUns 5.0 programs, three environment variables must be set. If you install the NCTUns 5.0 package into the default /usr/local/nctuns directory, in the following description ${where-you-install} should be replaced with /usr/local/nctuns. NCTUNSHOME should be set to ${where-you-install}, NCTUNS_TOOLS should be set to ${where-you-install}/tools, and NCTUNS_BIN should be set to ${where-you-install/bin E.g., if you use tcsh, you can add the following lines to your .cshrc shell configuration file setenv NCTUNSHOME ${where-you-install} setenv NCTUNS_TOOLS ${where-you-install}/tools setenv NCTUNS_BIN ${where-you-install}/bin E.g., if you use bash, you can add the following lines to your .bashrc shell file export NCTUNSHOME=${where-you-install} export NCTUNS_TOOLS=${where-you-install}/tools export NCTUNS_BIN=${where-you-install}/bin You can also use the ${where-you-install}/etc/nctuns.{csh|bash} to help you set all environment variables E.g., if you use tcsh, you can add the following lines to your .cshrc shell configuration file source ${where-you-install}/etc/nctuns.csh E.g., if you use bash, you can add the following lines to your .bashrc shell configuration file source ${where-you-install}/etc/nctuns.bash If you do not add these commands to your .cshrc or .bashrc file, you will need to manually set these variables in each opened xterm window before executing a NCTUns 5.0 program. )