The operation reliability of the power transformer as the key equipment in electrical power systems , influences operation security of electrical power systems directly . the components and contents of gases dissolved in transformer oil can be used to reflects internal insulation faults of operating transformer . in order to overcome the errors caused by complex handling procedure and man - made factors using general chromatogram analysis method , author brings forward an on - line detecting of gases dissolved in transformer oil by using macromolecule polymer to separate oil and gases automatically and an information fusion technology of multi - sensors ; at the same time , in order to improve the accuracy and reliability , author uses neural networks to diagnose transformer faults 变压器作为电力系统的枢纽设备，其运行可靠性直接影响电力系统的安全运行；变压器油中溶解气体的成分和含量能有效体现运行变压器内部的绝缘故障情况，为解决常规色谱分析中复杂的操作程序和由于人为因素引起的较大的误差，论文提出了应用高分子聚合膜实现变压器油气自动分离、多传感信息融合技术智能检测多种气体成分的变压器油中溶解气体在线监测技术，应用神经网络智能诊断方法实施故障诊断，提高变压器故障诊断的准确性和可靠性。
In the off - line learning , ga is used to optimize the global parameters of member functions . during the on - line learning , the algorithm inherited from the principle of back propagation ( bp ) is used to train the local parameters of network weights . applying it into position servo system , satisfied effect is obtained 在误差较小时，采用基于遗传算法的模糊神经网络智能控制，基于遗传算法的模糊神经网络智能控制器，它用遗传算法优化具有全局性的隶属度函数参数，用bp算法在线调节具有局部性的网络权值参数。
Furthermore intelligent shift control thinking and method are introduced detailed , and the intelligence control of fuzzy neural network is led to construct the model of the speed control of engine , the model of decision of shift , the model of clutch joint velocity and the model of the whole amt gearshift , simulation is carried out 详细地描述了智能换档控制的思路和方法，引入模糊神经元网络智能控制技术，构造发动机转速控制模型、档位决策模型、离合器接合速度控制模型及amt换档系统模型，并进行仿真模拟。
