公开(公告)号：WO2013110324A1

公开(公告)日：2013-08-01

申请号：PCT/EP2012/051062

申请日：2012-01-24

Applicant: ABB TECHNOLOGY AG ,  MERCANGOEZ, Mehmet ,  PARESCHI, Diego ,  BØHAGEN, Bjørnar ,  BESSELMANN, Thomas

Inventor： MERCANGOEZ, Mehmet ,  PARESCHI, Diego ,  BØHAGEN, Bjørnar ,  BESSELMANN, Thomas

IPC: F04D27/00

CPC classification number: F04D27/0207 ,  F04D27/001 ,  F04D27/0261 ,  F04D27/0284

Abstract: Compressor control including both surge detection and surge prevention. In one aspect of the invention, upon surge detection the rotating speed of the compressor is modified before a relief valve is opened. In a second aspect of the invention, a time-dependent operating map and compressor model is used for automatic control. The model includes inputs such as rotation speed and valve position. At each real time instant, the time-dependent map and model is linearised around the current state of the compressor. Conditions such as a minimum distance to surge line, desired output pressure, desired massflow, maximum rotation speed, etc... are expressed as mathematical constraints on the variables at future time steps. The linearised model is solved for the inputs at the current time step.

Abstract translation: 压缩机控制包括浪涌检测和浪涌预防。 在本发明的一个方面中，在浪涌检测之后，压缩机的旋转速度在安全阀打开之前被改变。 在本发明的第二方面中，使用时间依赖的操作图和压缩机模型来进行自动控制。 该模型包括旋转速度和阀门位置等输入。 在每个实时瞬间，时间依赖的地图和模型围绕压缩机的当前状态进行线性化。 诸如到喘振线的最小距离，期望的输出压力，期望的质量流量，最大转速等的条件被表示为在将来的时间步长对变量的数学约束。 线性化模型解决了当前时间步长的输入。

公开(公告)号：WO2015104281A1

公开(公告)日：2015-07-16

申请号：PCT/EP2015/050148

申请日：2015-01-07

Applicant: ABB TECHNOLOGY AG

Inventor： DOMINGUEZ, Luis ,  MERCANGOEZ, Mehmet ,  CORTINOVIS, Andrea ,  ZEYDAN, Burak

IPC: G01W1/10

CPC classification number: G01W1/10 ,  F24S2201/00 ,  G01W1/12

Abstract: The present invention is concerned with the operation of solar power plants such as Photo Voltaic (PV) or Concentrated Solar thermal Power (CSP) plants. The invention involves recursive updating of estimated parameters pertaining to pre-defined and physically motivated cloud evolution models that are instantiated and assigned to clouds identified in images taken by a ground-based visible-light camera. Identification of clouds, which includes assigning individual pixels of an image to one or several identified clouds, and selection of matching cloud models invoke some elaborate complexity beyond conventional pixel-wise image processing. This ultimately allows to predict a cloud-wise, as opposed to a pixel-wise, evolution based on cloud evolution models with a limited number of estimated parameters.

Abstract translation: 本发明涉及太阳能发电厂如光伏（PV）或集中太阳能热电（CSP）工厂的运行。 本发明涉及递归更新关于预定义和物理动机的云演化模型的估计参数，其被实例化并分配给由基于地面的可见光相机拍摄的图像中识别的云。 识别云，其包括将图像的各个像素分配给一个或多个识别的云，以及匹配的云模型的选择引起了超越常规像素图像处理的一些复杂的复杂性。 这最终允许基于具有有限数量的估计参数的云演化模型来预测云计算，而不是基于像素的演进。