中文摘要：

采用递阶遗传算法学习径向基函数神经网络的结构和参数，实现了某扫雷犁电液伺服系统的神经网络精确建模。介绍了某扫雷犁电液伺服系统的工作原理，详细论述了基于递阶遗传算法的径向基函数神经网络设计方法，即采用控制基因确定隐节点数目，采用参数基因辨识隐节点中心参数，利用最小二乘法计算连接权值。最后采用该方法对某扫雷犁电液伺服系统进行了建模研究，并与其它建模方法进行了比较，实验结果验证了该方法的有效性。

英文摘要：

A radial basis function neural network modelling method based on hierarchical genetic algorithm was proposed and applied to the hydraulic system of a certain type of mine sweeping plough. The operation principle of the hydraulic system was introduced. The neural network modelling technique using hierarchical genetic algorithm was detailed. The number of hidden units was determined by the control genes and the parameters of hidden centers were identified by the parameter genes, and the weighs were calculated using the least square method. The proposed method was applied to the electrohydraulic system, the results and comparisons with other modelling techniques clearly show its validity.