基于神经网络和遗传算法的机泵一体化优化方法的研究

发布时间：2018-05-12 10:42

本文选题：BP神经网络 + 遗传算法　；参考：《安徽大学》2015年硕士论文

【摘要】：为解决实际工程应用中泵水力模型与电机电磁计算模型的结合问题,本文采用BP神经网络和遗传算法(GA-BP)对水泵的水力性能进行拟合建模,并结合电机电磁计算模型建立机泵一体模型。本文的主要工作如下：1.在BP神经网络和遗传算法的基本原理基础上,研究使用遗传算法对泵水力模型中的BP神经网络结构、权值、阀值的进行优化；并运用C语言实现基于GA-BP算法的泵水力性能的BP神经网络建模；2.将GA-BP算法的建模结果和基本BP神经网络繁荣建模结果进行对比及分析,从精度、速度、网络规模、泛化能力等方面验证了GA-BP算法建模的有效性；3.电机电磁计算模型与泵水力模型以接口变量的形式连接成泵机一体化模型,并对基本BP神经网络和GA-BP算法的运行结果进行了比较分析。论文的研究工作得到国家科技支撑计划课题“矿山用大型主排水泵电机系统节能技术的研究与应用”(2013BAF01B01)的资助。
[Abstract]:In order to solve the problem of the combination of pump hydraulic model and motor electromagnetic calculation model in practical engineering application, this paper uses BP neural network and genetic algorithm to simulate the hydraulic performance of the pump. Combined with the electromagnetic calculation model of motor, the integral model of motor and pump is established. The main work of this paper is as follows: 1. Based on the basic principle of BP neural network and genetic algorithm, the optimization of BP neural network structure, weight and threshold in pump hydraulic model by genetic algorithm is studied. The BP neural network model of pump hydraulic performance based on GA-BP algorithm is realized by C language. The modeling results of GA-BP algorithm and basic BP neural network are compared and analyzed. The validity of GA-BP algorithm modeling is verified from the aspects of precision, speed, network scale and generalization ability. The electromagnetic calculation model of motor and the hydraulic model of pump are connected in the form of interface variables to form the integrated model of pump machine. The results of basic BP neural network and GA-BP algorithm are compared and analyzed. The research work of this paper is supported by the national science and technology support project, "Research and Application of Energy-saving Technology for large Mine main drainage pump Motor system" (2013BAF01B01).
【学位授予单位】：安徽大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD636;TP18

【参考文献】