查询词典 neural network

On the basis of variation characteristics of ship speed error model, a ship speed regulator using the on-line adaptive error model is proposed in this paper. The control scheme is based on back propagation multi-layer perceptrons adaptive neural network in which a modified hyperbolic tangent function is used as the activation function of an auto-tuning neuron and is updated by the steepest descent method.

在分析船舶航速误差模型变化特性的基础上，应用激励函数的自适应神经网络和最陡下降法的非线性优化技术，提出了自适应学习误差模型控制方法，应用于船舶柴油机调速器控制系统。

In this study, the following studies have been carried out(1)Test on a typical MR damper employed in this study was carried out, Neural network model based a direct model and a inverse model were established. The performance of the models were cpmpared with the Bingham model and the hyperbolic tangent model. Results show that ANN model can describe the mechanical behavior of MR damper perfectly as a nonparmetric model.

本文主要进行了以下三个方面的工作：(1)对一种典型的磁流变阻尼器进行了性能试验，利用实验实测数据建立了阻尼器的Bingham模型、双曲正切模型两种参数模型以及基于神经网络的正向和逆向模型，对三种计算模型进行了仿真分析，并与试验结果进行对比研究，结果表明，神经网络模型比两种参数模型能更好地模拟阻尼器的动态性能。

Based on analyzing the multilayer perceptron neural network,it induces the error backpropagation algorithm while taking the hyperboloid tangential function as nonlinear activation function.

在分析多层感知器神经网络的基础上，推导了双曲正切函数为激励函数的误差后向传播算法。

An unsupervised hypersphere neural network that is used for pattern clustering is presented.

摘要提出和实现了用于模式聚类的无监督模糊超球神经网络。

The simulation of clustering demonstrates the superiority of the fuzzy hypersphere clustering neural network.

模式聚类的数值仿真解释了模糊超球聚类神经网络的优越性能。

It is using Neural Network that the hypha density, total sugar density and product density, is estimated by pH and dissolved oxygen.

本文采用神经网络方法，通过pH、溶氧这两个参数来估计菌丝浓度、基质浓度和产物浓度。

The simulation results also indicate that the neural network predictive can effectively solve the problem induced by hysteresis.

本文采用的神经网络预测的方法能有效的改善磁滞效应所产生的问题。

In this dissertation, some problems based on identification and control of neural network have been studied.

基于此，本文对基于神经网络辨识与控制的若干问题进行了较深入地研究。

In the second part, we studied blind source separation in ill conditioned cases, obtained the sufficient and necessary conditions on solvability, and presented the neural network model and corresponding adaptive algorithm. We also discussed the blind digital source separation, established the blind separation and extration models and algorithms. These algorithms can be used to deal with high noise and ill-conditioned cases.

在第二部分，我们研究了奇异混合情形下信号盲分离问题，给出了神经网络模型及相应的自适应算法；我们也研究了数字信号盲分离问题，得到了有关可解性的充要条件，提出了分别处理低噪声问题和高噪声问题的算法，这些算法对感知信号个数少于源信号个数等病态问题也适用。

The technology, illation rule and parameter of neural network are supposed to be held in database and make a delicate database system.

知识和推理规则、神经网络参数都必须由数据库保存，使数据库系统实现知识／数据的一体化存储与管理。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。