退火算法在电子元件位置优化上的应用

Application of Annealing Algorithm in Optimal Placement of Electronic Components

一块电路板上有电子元件，它们不仅有热源，而且相互作用，另外与空气还交换热．对这类问题建立线性数学模型，从而算出温度稳态分布．因为许多集成电路系统的可靠性很大程度上取决于温度，所以温度的模拟显得很重要．把温度模型与退火算法相结合可以获得元件最接近优化的布局，并用一个例子加以说明，通过优化能降低板子最高温度，从而提高系统的可靠性．

Consider a board containing electronic components which (1) act as sources of heat, (2) interact thermally with their neighbors and (3) exchange heat with an air current. First, we construct a linear model for such problems with constant ambient temperature, use finite difference to solve this model and can specify the steady state temperature distribution explicitly in terms of problem data. The importance of simulating the temperature of electronic boards lies in the fact that the reliability of many IC system is temperature dependent. Consequently we use our temperature model in conjunction with the annealing algorithm to obtain near optimal configurations of the components. Finally, we use the annealing algorithm together with our main results to find the optimal placement of the IC components in an example. Optimization is interpreted as minimizing the maximum board temperature and increasing the reliability of the system.