论文部分内容阅读
在现有的遗传算法的基础上，采用面向对象技术设计了面向对象的遗传算法，建立了遗传算法的类层次。这种方法改变了在传统的遗传算法中各个函数之间只有参数的传递，而没有代码的继承性的状况从概念上提高了软件的可重用性。该方法在人工神经网络的辅助设计问题中的应用表明，这一算法由于采用面向对象的分析与设计方法，从而具有比传统的遗传算法更好的通用性，用户可以更方便地设计和实现自己的编码方案和遗传算子，大大提高了软件的可重用性。
Based on the existing genetic algorithms, an object-oriented genetic algorithm is designed using object-oriented technology and a class hierarchy of genetic algorithms is established. This method changes the transmission of parameters between functions in a conventional genetic algorithm, and the absence of inheritance of code conceptually improves software reusability. The application of this method in the auxiliary design of artificial neural networks shows that this algorithm has better generality than the traditional genetic algorithm due to the adoption of the object-oriented analysis and design method, and the user can design and realize himself more conveniently The coding scheme and genetic operators greatly improve the software reusability.