房子 跳舞水 在澳门

PEN

http://deltamotion.com/peter/Videos/HODW%20v6%2020181227.mp4
我们终于发布了一个视频，介绍我们如何控制澳门舞池水平的平台。我们获得了制作视频的权限和视频。该视频由论坛成员deltamotion-usa制作。提供一个好节目的努力是惊人的。观众没有看到太多。

我们的经销商Comoso提供液压系统。 Comoso使用Parker和Delta运动控制器。 Fischer Technical做了编程。 Fischer Technical是主承包商。他们使用Rockwell Contrologix PLC对11个RMC150进行排序。有53个液压缸可以同步。一切都必须精确安全。

平台非常庞大而且沉重。液压缸长9米，使用很长的Tempsonic传感器。气缸和负载的固有频率非常低。我几年前做过计算。我不知道的是阀门和气缸之间会有很多软管。这进一步降低了固有频率。

我们的工程师前往澳门提供建议。

接近结束时，你可以看到我在2013年做演讲.delmotion-usa正在为我翻译。杨世祥坐在前面，有疑问。你可以看到杨世祥的脑袋后面。

你可以看到我手腕骨折了。几天前我在乒乓球比赛中摔倒了，但我仍然前往参加我的演讲。 deltamotion-usa我喜欢打乒乓球。乒乓球是另一种运动控制形式。我们喜欢在浦东的源深体育中心玩。

PEN

我有一个.ppt文件的链接，该文件是5年前在上海PTC2013上用于演示的。
http://deltamotion.com/peter/ppt/DELTA_PTC-Asia_En-Ch_20131018.ppt
视频部分现在是独立的。这个视频加快了。
http://deltamotion.com/peter/Videos/The%20house%20of%20dancing%20water%20-%20lift%20cues%20timelapes.mp4

我最初的担忧是
1.低固有频率。
2.改变阻尼系数。
3.长的Temposonic棒。

RMC150有一个内置的模拟器。我获得了质量和圆柱尺寸。我使用此信息来估计固有频率，并在作为RMC150一部分的模拟器中使用这些值。我确定RMC150能够在规格范围内控制最重的平台。

长的Temposonic棒是一个问题。最初有人承诺RMC150会每2毫秒获得一次位置反馈，但这违背了杆的物理特性。由于气缸长9米，而Temposonic杆的常数约为2.8米/秒，因此从Tenmposonic杆上取回位置需要3毫秒。这意味着RMC150上的扫描时间必须增加到4毫秒。我觉得这样做很安全，因为平台的固有频率很低。

定位必须精确，以便演员不会偶然发现平台之间的差异。由于平台可以快速而精确地移动，因此它允许节目主管更多地选择如何移动风景。此外，即使平台上的风景发生变化，定位也必须精确。

调整需要我们相对较新的高级调整技术。带有前馈的PID不能很好地工作，因为固有频率和阻尼系数很低。

视频或图片中未明确的是控件位于水箱下方的房间内。必须支持水和平台的所有重量。气缸必须通过下方的阀门，控制器和动力装置穿过油箱底板。下面甚至有更多的管道减少了自然频率比原先估计的更多。

最后，多年后，我们制作了这个系统，可用于培训工程师用于控制低固有频率系统的技术。
http://deltamotion.com/peter/Videos/NF-FOA.mp4
该技术有效。

have a link to a .ppt file that was used for the presentation 5 years ago at PTC2013, Shanghai.
http://deltamotion.com/peter/ppt/DELTA_PTC-Asia_En-Ch_20131018.ppt
The video part is separate now.  This video is sped up.
http://deltamotion.com/peter/Videos/The%20house%20of%20dancing%20water%20-%20lift%20cues%20timelapes.mp4

My initial concerns were
1. low natural frequency.
2. changing damping factor.
3. the long Temposonic rods.

The RMC150 has a built in simulator.  I was provided with the mass and cylinder dimensions.  I used this information to estimate the natural frequency and used these values in the simulator that is part of the RMC150.  I determined that the RMC150 would be able to control the heaviest platform within specifications.

The long Temposonic rods were a problem.  Originally were were promised that the RMC150 would get position feedback every 2 milliseconds but this defies the physics of the rod.  Since the cylinders were 9 meters long and the Temposonic rod have a constant of about 2.8 m/ms,  it was going to take over 3 milliseconds to get an positon back from the Tenmposonic rod.  This meant that the scan time on the RMC150 had to be increased to 4 milliseconds.   I felt it was safe to do because the natural frequency of the platforms was so low.

The positioning had to be precise so the actors would not stumble on the differences between the platforms.  Since the platforms could be moved quickly and precisely, it allowed the show director more options on how the scenery can be moved.  Also, the positioning had to be precise even though the scenery on the platforms change.

The tuning required our relatively new advanced tuning techniques.  PID with feed forwards would not work well because the natural frequency and damping factor are so low.

What isn't made clear in the videos or pictures is that the controls were in a room below the tank.  All the weight of the water and platforms had to be supported.  The cylinder had to stick up through the tank floor with the valves, controllers and power units below.  There was even more piping below that reduced the natural frequency more than originally estimated.

Finally, many years later we made this system that can be used to train engineers on the techniques used to control low natural frequency systems.
http://deltamotion.com/peter/Videos/NF-FOA.mp4
The technology works.