The self - tuning fuzzy control algorithm including a back - propagation - based fuzzy controller and the gradient descent method is applied to reduce the pitch motion of the swath ship with control stabilizing fins 控制模式为一自调模糊控制，包括应用一类神经倒传递观念的模糊控制器及梯度下降法来减少纵摇运动。
Secondly , adaptive - network based fuzzy inference systems ( anfis ) turn membership functions in antecedent part and parameters in consequent par of inference rules by means of least - square methods and gradient descent methods 与普通的模糊神经网络相比，这种模糊神经网络收敛速度快，建模精度高。第三，混合模糊神经网络建模。
At present , during the course of bp neural network ' s learning and training , we often adopt the algorithm of back - error propagation , which is based on global error function ' s gradient descent 特别是bp网络近年来广泛应用于模式识别、预测评估等领域，并取得良好的效果。目前bp网络采用误差逆传播算法学习训练神经网络，该算法是基于网络误差函数梯度下降的。
At present , during the course of bp neural network ' s learning and training , we often adopt the algorithm of back - error propagation , which is based on global error function ' s gradient descent , whose essence is a point - to - point search algorithm 目前bp神经网络采用误差逆传播算法学习训练神经网络，该算法是基于网络误差函数梯度下降的，其本质是点到点的搜索方法。
The gaussian function set depends on its centroid and width , therefore , we derive new supervised gradient descent algorithms to tune the parameters of " if - part " of gaussian functions . it also tunes the " then - part " of gaussian function 高斯型函数是由它的中心和宽度两个参数决定的，所以，我们给出新的有监督的梯度下降算法来调整如果部分高斯函数与则部分的高斯函数的中心和宽度。
It uses genetic algorithm for optimizing intrinsic parameters and gradient descent method for optimizing extrinsic parameters . compared with classical techniques , the proposed technique is easier , more accuracy and robust , furthermore , it needn ’ t estimate initial parameters 该标定方法内部参数的优化使用遗传算法,外部参数的优化使用梯度下降法,它具有原理简单、不需要估计初始值、鲁棒性强、标定参数更准确的优点。
Secondly , by incorporating the temporal difference prediction technique with the genetic algorithm , a novel hybrid genetic neural networks for controlling nonlinear chaotic system based on the scheme of small perturbations and the use of gradient descent learning method is presented ( known as hygann strategy ) 2 .研究了一种将暂态误差预测技术、小扰动控制技术、梯度下降法和遗传算法( ga )融合起来控制非线性混沌系统的复合遗传神经网络方法(简称hygann法) 。
Procreant knowledge expression and forward inference engine are adopted in the method of fault diagnosis based on expert system theory . in the fault diagnosis applying neural network theory , six kinds of improved arithmetic of back - propagation arithmetic , including gradient descent with momentum , variable learning rate back - propagation , resilient back - propagation , quasi - newton , levenberg - marquardt and conjugate gradient , are applied to diagnose the faults of electric load manage center and solid state power controller . different diagnostic results gotten by simulation are compared at last 在基于专家系统的故障诊断方法中，采用了产生式知识表达和正向推理机制；在基于神经网络的故障诊断方法中，则分别采用了bp神经网络的附加动量法、自适应学习速率、弹性bp算法、拟牛顿法、共轭梯度法和levenberg - marquardt法对电气负载管理中心和固态功率控制器的故障进行诊断，并对由仿真得到的不同诊断结果进行比较。
Because of its complicated dynamics property , chaotic neural network has recently attracted some attention . unlike the gradient descent neural network , the chaotic neural networks has more complex dynamics property , and diversified attractor exist . it is just the dynamics that make it possible for the network to be a technology with abroad application foreground for information processing and optimization computation 由于具有混沌特性的人工神经网络具有十分复杂的动力学特性，获得了广泛的研究，不同于仅具有梯度下降特性的常规神经网络，具有混沌特性的神经网络具有更加丰富的和远平衡点的动力学特性同时存在各种吸引子。
So the paper combined the fuzzy logic control and recurrent neural network , and the recurrent fuzzy neural network ( rfnn ) controller is introduced into the speedsensorless vector control system . moreover , an online parameter training methodology , which is derived from the lyapunov stability theorem and gradient descent method is proposed to increase the learning capability of the rfnn . the rfnn controller has a better performance than the pi controller system ; the effectiveness of the proposed control scheme is verified by simulation results 因此论文将模糊控制和具有优越动态性能的递归神经网络结合起来，取长补短，提出了一种递归模糊神经网络控制方法，利用神经网络来实现模糊推理，可动态的调整隶属函数的形状、位置以及神经网络递归权值，并对其与pi控制器的交流调速控制系统进行了仿真比较，仿真结果表明与普通的pi控制器相比较，递归模糊神经网络控制器有较好的动态性能，控制器的收敛速度快、静差小，系统在遇到参数发生变化和外部不确定性问题时鲁棒性、抗扰动性有明显的提高。
