2D Regions and Faces

 

2D Regions and Faces

CFX-Mesh will often ask you to select one or more 2D Regions for the location of a meshing feature. Where two Solid Bodies meet at a common face, there is just one face present in the geometry; however, there are two 2D Regions, as shown in the example below.

If your model has only one solid, then each face is also a 2D Region, and vice versa. If your model has more than one solid, then all of the external faces will also be 2D Regions. However, two 2D Regions will exist for every shared (internal) face (assuming that you have combined the solids into the same part). Each meshing feature that requires you to specify a location has its own rules about whether or not you can set different properties for the two 2D Regions which make up a shared face.

  • Composite 2D Region: You may include any combination of 2D Regions in a Composite 2D Region (including just one or both sides of a shared face). However, if you are going to use a Composite Region to specify the location for any other meshing feature, then you must make sure that it only includes 2D Regions which satisfy the requirements of the appropriate feature.

  • Face Spacing: It is important that you do not try to apply different Face Spacings to 2D Regions which are the two sides of a common face, since the surface mesh is generated on the common face, not on the 2D Regions separately. It is acceptable to include a 2D Region which forms one half of a shared face in a Face Spacing without including the other side of the shared face, but only if the 2D Region which forms the other side is not included in any other Face Spacing (other than the Default Face Spacing).

  • Inflated Boundary: When creating an Inflated Boundary, you can have different settings for the two different 2D Regions which make up a common face, i.e., you can apply Inflation to one 2D Region and not the other, or to both but with different settings. The example below shows the difference in one particular case.

  • Preview Group: A Preview Group can contain any combination of 2D Regions. It does not matter if 2D Regions are contained in more than one Preview Group, nor if a 2D Region is included which forms one half of a shared face, but the other half is not included.

  • Virtual Faces: A Virtual Face cannot include any 2D Region which forms one half of a shared face.

If you try to pick a 2D Region which is half of a shared face, then the Selection Rectangles will appear. Each shared face will be represented by two rectangles which are attached to each other, one for each side or 2D Region, as shown below. You can use these rectangles to select individual 2D Regions easily and accurately.

CFX-Mesh will not allow you to select locations for meshing features which break the rules given for each feature.

使用优化算法,以优化VMD算法的惩罚因子惩罚因子 (α) 和分解层数 (K)。 1、将量子粒子群优化(QPSO)算法与变分模态分解(VMD)算法结合 VMD算法背景: VMD算法是一种自适应信号分解算法,主要用于分解信号为不同频率带宽的模态。 VMD的关键参数包括: 惩罚因子 α:控制带宽的限制。 分解层数 K:决定分解出的模态数。 QPSO算法背景: 量子粒子群优化(QPSO)是一种基于粒子群优化(PSO)的一种改进算法,通过量子行为模型增强全局搜索能力。 QPSO通过粒子的量子行为使其在搜索空间中不受位置限制,从而提高算法的收敛速度与全局优化能力。 任务: 使用QPSO优化VMD中的惩罚因子 α 和分解层数 K,以获得信号分解的最佳效果。 计划: 定义适应度函数:适应度函数根据VMD分解的效果来定义,通常使用重构信号的误差(例如均方误差、交叉熵等)来衡量分解的质量。 初始化QPSO粒子:定义粒子的位置和速度,表示 α 和 K 两个参数。初始化时需要在一个合理的范围内为每个粒子分配初始位置。 执行VMD分解:对每一组 α 和 K 参数,运行VMD算法分解信号。 更新QPSO粒子:使用QPSO算法更新粒子的状态,根据适应度函数调整粒子的搜索方向和位置。 迭代求解:重复QPSO的粒子更新步骤,直到满足终止条件(如适应度函数达到设定阈值,或最大迭代次数)。 输出优化结果:最终,QPSO算法会返回一个优化的 α 和 K,从而使VMD分解效果最佳。 2、将极光粒子(PLO)算法与变分模态分解(VMD)算法结合 PLO的优点与适用性 强大的全局搜索能力:PLO通过模拟极光粒子的运动,能够更高效地探索复杂的多峰优化问题,避免陷入局部最优。 鲁棒性强:PLO在面对高维、多模态问题时有较好的适应性,因此适合海上风电时间序列这种非线性、多噪声的数据。 应用场景:PLO适合用于优化VMD参数(α 和 K),并将其用于风电时间序列的预测任务。 进一步优化的建议 a. 实现更细致的PLO更新策略,优化极光粒子的运动模型。 b. 将PLO优化后的VMD应用于真实的海上风电数据,结合LSTM或XGBoost等模型进行风电功率预测。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值