BBL布局的均场退火方法

A Mean Field Annealing Approach for BBL Placement

均场退火方法既可以看作是一种新的神经网络计算模型，又可视为是对模拟退火的重大改进．提出了一个基于均场退火方法的任意单元布局算法，用一个三维二值换位矩阵将问题映射为神经网络，建立包含重叠约束和优化目标的能量函数，再用均场退火方程迭代求解．每个单元只能放置在布局平面一个位置上的约束。

The mean field annealing approach is a new neural network model, which improves simulated annealing approach greatly. In this paper, a mean field annealing approach for general cell placement is proposed. In the algorithm, a three dimensional permute matrix of binary variables is used to map the problem to the neural network, the energy function including object item and overlap constrained item is presented, and then iteration procedure is put into practice with the mean annealing equation. Normalization of neurons proves that one cell can only be assigned to one position in the placement grid.