混合优化自动剩余静校正方法研究

剩余静校正量计算本质上是一个非线性优化问题。针对单独运用遗传算法或模拟退火算法求解剩余静校正计算效率低、精度不高等缺点，本文提出了混合全局优化的自动剩余静校正方法，即利用两者的优点（遗传算法具有较强的把握搜索过程总体的能力，模拟退火法具有较强的局部搜索能力）开发的一种优化算法。文中对混合全局优化自动剩余静校正的处理流程和实现技术作了深入研究。从模型数据试算和我国西部某地区实际资料的处理结果可以看到，混合优化技术无论在运算速度还是在求解精度上都优于其他方法。

Study of Method of Hybrid Optimum Automatic Residual Static Corrections

The nature of computation of residual statics is non linear optimum problem. Faced with the disadvantages of slower computational speed and lower precision that are characters using genetic algorithm or simulated annealing algorithm alone to compute residual statics, the paper presented an automatic residual static corrections characterized by hybrid global optimization,i, e. an optimized algorithm developed by using the both advantages ( genetic algorithm has strong ability to search globally and simulated annealing algorithm has strong ability to search locally). The paper deeply studied processing flow and implemented techniques of hybrid global optimum automatic residual static corrections. It is seen from test computation of model data and real data in some areas of the west of China that the hybrid optimum technique is superior to other methods both in computational speed and in precision of solution.