中文摘要：

针对测试选择优化这一NP难解问题，文中利用改进的遗传算法对其进行优化。算法以遗传算法为主流程．利用混沌现象不重复遍历的特点优化生成初始种群。然后对每次迭代中的个体以一定的概率进行混沌优化。最后，以超外差接受器为例，对算法的有效性进行了验证。事实证明，该算法能够较快地搜索到优化问题的最优解，验证了混沌遗传算法对测试选择优化问题的有效性。

英文摘要：

The improved genetic algorithm （GA） was wsed to solve test selection optimization which is a NP-hard problem. This improved algorithm takes GA as the main flow, and uses the character of traversing without repeat of chaos to optimize the initial group and then optimizes each individual with certain probability. Finally, it takes super-heterodyne aeceptor as an example to validate effectiveness of this improved algorithm. It shows that this algorithm can get globally optimal solution quickly; this proves that chaos genetic algorithm can solve problem of test selection optimization effectively.