遗传算法在路面结构模量反算中的应用

The Application of Genetic Algorithms in the Backcalculation of Pavement Structure Moduli

利用遗传算法反算路面结构层弹性模量可以避免一般反算方法受初值影响的缺陷。探讨了遗传算法在路面结构模量反算中的应用 ,并开发了反分析程序BACKGA ;计算结果表明 ,该方法计算精度很好 ,但计算时间较长。为提高程序计算效率 ,利用神经网络计算路面结构弯沉值 ,并将其用于反分析程序BACKGA中 。

The backcalculation of pavement layer moduli through genetic algorithms can eliminate the limitation of current algorithms that the result is influenced by initial value. In this paper, the authors discuss the application of genetic algorithms in the backcalculation of pavement layer moduli. The program BACKGA is also presented to apply to the backcalculation of moduli. The result of computation shows that the precision of the result is very good, although the program has a relatively long computation time. Therefore, neural network theory is adopted to compute the deflection of pavement structure layers. The computation speed can be accelerated greatly by using neural network.