基于ROLS算法的RBF神经网络燃料电池电特性建模

提出了一种基于ROLS算法的RBF神经网络辨识建立直接甲醇燃料电池(DMFC)电特性模型的新方法。以电池的工作温度为输入量,电池的电压/电流密度为输出量,利用1200组实验数据作为训练和测试样本,建立了在不同工作温度下,电池的电压/电流密度动态响应模型。仿真结果表明采用RBF神经网络辨识建模的方法是有效的,建立的模型精度较高。

自主学习模糊神经网络PID加热炉炉压控制

炉膛压力的稳定将直接影响加热炉的加热效率和加热质量。但由于加热材料的周期性进出,对炉压会产生较大的干扰,采用传统的PID控制已经不能满足实时控制要求。现提出一种自主学习模糊RBF神经网络辨识PID在线实时控制炉压,控制效果更加稳定。

Model Identification of Water Purification Systems Using RBF Neural Network

Aim The RFB (radial hats function) netal network was studied for the model indentificaiton of an ozonation/BAC system. Methods The optimal ozone's dosage and the remain time in carbon tower were analyzed to build the neural network model by which the expected outflow CODM can be acquired under the inflow CODM condition. Results The improved self-organized learning algorithm can assign the centers into appropriate places , and the RBF network's outputs at the sample points fit the experimental data very well. Conclusion The model of ozonation /BAC system based on the RBF network am describe the relationshipamong various factors correctly, a new prouding approach tO the wate purification process is provided.

一种基于PF-RBF的ANNPID参数自整定方法

针对实际控制过程中控制对象难以建立精确模型和由于非线性非高斯噪声干扰导致控制效果难以达到预期的问题,提出了一种基于粒子滤波和RBF神经网络辨识(RBF Neural Network and Particle Filter Algrothm,PF-RBF)的单神经元PID参数自整定方法。通过PF和RBF系统辨识得到精确的系统Jacobian信息,解决神经网络PID控制由于Jacobian信息未知导致的近似计算不精确问题。仿真实例表明,相对于无PF滤波的RBF辨识控制系统,该方法能改善控制系统的性能指标和抗干扰能力,对实际控制过程具有一定的指导意义。