基于改进BP神经网络算法的微波加热智能控制系统的研究

发布时间：2018-07-03 13:36

本文选题：微波 + 温度控制　；参考：《昆明理工大学》2014年硕士论文

【摘要】：微波作为一种新型加热技术,以其高效、省时、加热均匀以及无污染等特点,在工业、农业以及医疗等领域体现出巨大的优势和广阔的前景。因此,随着工业微波加热技术不断地创新与发展,微波输出功率的精确及智能控制将成为微波加热技术发展的关键。本文设计了一种基于改进共轭梯度的BP神经网络PID算法的工业微波加热自动控制系统。文章主要阐述了磁控管作为微波源工作的基本原理、微波加热的基本特点、温度检测与控制的硬件电路以及相应的智能控制算法。本文的主要工作如下： 1.概述了微波加热的原理和基本特点,介绍了微波加热系统的结构,详细说明了磁控管的基本结构以及工作原理。通过理论分析,对微波传输装置和加热腔进行了研究和设计。 2.针对微波加热环境,分析了几种类型温度传感器的优缺点,对热电偶温度传感器进行了改进分析。完成了以DSP为核心处理器的硬件电路设计,包括采集、监测和控制三部分。温度采集电路由K型热电偶和MAX6675集成芯片组成；温度监测部分设计了LCD液晶显示和通过MAX232串行接口通信的上位机软件实现对温度变化的实时监测；控制部分利用PWM脉冲调制技术实现对磁控管输出功率的调节,从而实现温度的智能控制。 3.针对温度非线性、滞后性和时变性等特点,采用改进共轭梯度的BP神经网络PID算法。利用MATLAB软件分别对常规的增量式PID算法、BP神经网络PID算法和改进共轭梯度的BP神经网络PID算法进行仿真研究,从理论上证明了具有响应速度快、抗干扰性强以及控制精度高等特点的改进共轭梯度的BP神经网络PID算法在微波加热智能控制系统上应用的可行性。最后通过活性炭加热实验,验证了基于DSP和改进共轭梯度的BP神经网络PID算法的微波加热温度控制系统的实用性和有效性。
[Abstract]:As a new type of heating technology, microwave is characterized by its high efficiency, time saving, uniform heating and no pollution. It shows great advantages and broad prospects in the fields of industry, agriculture and medical treatment. Therefore, with the continuous innovation and development of industrial microwave heating technology, the precision and intelligent control of the output power of micro wave will become microwave heating. The key to the development of technology.
This paper designs an industrial microwave heating automatic control system based on the improved BP neural network PID algorithm based on the improved conjugate gradient. The article mainly expounds the basic principle of the magnetron as the microwave source, the basic characteristics of the microwave heating, the hardware circuit of the temperature detection and control and the intelligent control algorithm of the phase response. As follows:
1. the principle and basic characteristics of microwave heating are summarized. The structure of the microwave heating system is introduced. The basic structure and working principle of the magnetron are explained in detail. Through theoretical analysis, the microwave transmission device and the heating chamber are studied and designed.
2. according to the microwave heating environment, the advantages and disadvantages of several types of temperature sensors are analyzed, and the thermocouple temperature sensor is improved. The hardware circuit design with DSP as the core processor is completed, including three parts, acquisition, monitoring and control. The temperature acquisition circuit is composed of K thermocouple and MAX6675 integrated chip; temperature monitoring unit The LCD liquid crystal display and the host computer software that communicate through the MAX232 serial interface are designed to realize the real-time monitoring of the temperature change, and the control part uses the PWM pulse modulation technology to adjust the output power of the magnetron to realize the intelligent control of the temperature.
3. in view of the characteristics of temperature nonlinear, hysteresis and time-varying, the BP neural network PID algorithm with improved conjugate gradient is adopted. By using MATLAB software, the conventional incremental PID algorithm, BP neural network PID algorithm and the improved conjugate gradient BP neural network PID algorithm are simulated, and the fast response speed and dry resistance are proved theoretically. The feasibility of applying the conjugate gradient BP neural network PID algorithm to the microwave heating intelligent control system is improved. Finally, the practicability and effectiveness of the microwave heating temperature control system based on the BP neural network PID algorithm based on DSP and the improved conjugate gradient are verified by the activated carbon heating experiment.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM924.76

【参考文献】