论文部分内容阅读
文化算法是一种模拟文化进化过程的优化算法,它由基于个体和群体特性的信念空间和基于个体行为的种群空间组成,为进化搜索机制和知识存储的结合提供一个构架。建立基于生产过程输入输出数据的统计模型时,参数估计是其中的关键,文化算法为此提供了有效途径。本文研究用文化算法实现多变量优化的具体步骤、算法和关键环节的实施。建立裂解炉裂解深度的神经网络模型,并用文化算法优化网络参数,实验表明,文化算法比标准遗传算法搜索性能更优,搜索时间更快,同时得到了满意的裂解深度模型。
Cultural algorithm is a kind of optimization algorithm that simulates the process of cultural evolution. It consists of belief space based on individual and group characteristics and population space based on individual behavior, which provides a framework for the combination of evolutionary search mechanism and knowledge storage. When establishing a statistical model based on the input and output data of the production process, parameter estimation is one of the key points, and the cultural algorithm provides an effective way for this. This paper studies the concrete steps, algorithms and implementation of key steps to achieve multivariate optimization with cultural algorithms. The neural network model of pyrolysis depth of pyrolysis furnace was set up and the network parameters were optimized by using the cultural algorithm. Experiments show that the cultural algorithm has better performance than the standard genetic algorithm, the search time is faster, and the satisfactory pyrolysis depth model is obtained.