基于智能PID的小功率柴油发电机调速器的研制

发布时间：2018-06-21 10:50

本文选题：柴油发电机 + 数字电子调速器　；参考：《电子科技大学》2014年硕士论文

【摘要】：柴油发电机组被广泛应用在社会生产中,如机关单位、野外作业等领域的备用电源。发电机组的速度特性对供电质量影响较大,为了保证供电品质,柴油发电机组都需配备调速器,以确保负载变化时,柴油机转速能保持稳定。当前我国中小功率柴油发电机组基本依然采用模拟电子调速器,数字电子式的很少,且控制方法多为常规PID,难以满足复杂工况下的调速要求。鉴于此,本文展开了智能控制方面的研究,以期开发出性能可靠、通用性强的数字电子调速器。本文分析研究了柴油发电机的工作原理、结构组成及柴油机的工作特性,在此基础上,展开了柴油机调速系统的讨论,给出了调速系统的数学模型,提出了几种评判系统性能优劣的指标,如稳态波动率、瞬时调速率以及恢复时间等。控制算法方面,本文在常规数字PID基础上,引入了神经网络和模糊控制,在MATLAB下建立了单神经元PID和模糊自适应PID的控制系统模型,并分别进行了阶跃响应测试、抗干扰测试以及自适应性测试。仿真结果表明,基于智能控制的PID调速系统,具有超调量小、响应迅速及调速平稳的特点,系统动、静态品质得到显著改善;且控制对象参数变化时,系统适应性更强,鲁棒性更好,能较好满足柴油机多工况运行要求。考虑到模糊规则主要依据专家经验制定,系统性能受人为因素影响较大,而单神经元PID是根据控制对象的工况特性,在线修正连接权值、寻找PID参数最佳组合的,受外界影响较小,且算法简单、易于编程实现,故本文最终选择单神经元PID作为数字调速器的基本算法。电路设计方面,本文硬件电路包括:单片机最小系统、电源模块、转速信号整形电路、执行器驱动电路、电压电流采样电路以及其他辅助电路。其中,微处理器为16位的PIC24FJ64GA002型单片机,利用其高精度输入捕捉模块采集转速信号,输出16位分辨率的PWM,控制执行器驱动电路中三极管的通断,通过改变PWM占空比来调节执行器中的电流,进而调节柴油机供油量。软件方面,本文详细介绍了软件设计的基本思想,核心算法为单神经元PID。在启动阶段,本文设计了特殊的正弦式油量控制曲线,以实现烟量控制。此外,本文在康明斯220KW柴油发电机组上进行了配机实验,实验结果表明,基于单神经元PID的数字电子调速器灵敏度高,控制精度高,调速性能优良,满足发电机组的使用要求。
[Abstract]:Diesel generator sets are widely used in social production, such as government units, field operations and other areas of backup power supply. In order to ensure the quality of power supply, diesel generator sets need to be equipped with governor to ensure that the speed of diesel engine can keep stable when the load changes. At present, analog electronic governor is still used in medium and small power diesel generator sets in our country. The digital electronic governor is rarely used, and the control methods are mostly conventional PIDs, so it is difficult to meet the requirements of speed regulation under complex operating conditions. In view of this, the research of intelligent control is carried out in this paper, in order to develop a digital electronic governor with reliable performance and strong versatility. In this paper, the working principle, structure composition and working characteristics of diesel generator are analyzed and studied. On the basis of this, the discussion of diesel governing system is carried out, and the mathematical model of the speed regulation system is given. Several indexes to evaluate the performance of the system, such as steady state volatility, instantaneous tuning rate and recovery time, are proposed. In terms of control algorithm, this paper introduces neural network and fuzzy control on the basis of conventional digital pid, establishes the control system model of single neuron pid and fuzzy adaptive pid under MATLAB, and carries out step response test respectively. Anti-jamming test and self-adaptation test. The simulation results show that the pid speed control system based on intelligent control has the characteristics of small overshoot, quick response and stable speed regulation, and the dynamic and static quality of the system is improved significantly, and the system is more adaptable when the parameters of the control object change. The robustness of diesel engine is better and it can better meet the requirements of diesel engine running under multiple working conditions. Considering that the fuzzy rules are mainly based on expert experience, the system performance is greatly affected by human factors, and the single neuron pid is based on the operating conditions of the control object, the connection weight is online revised to find the best combination of pid parameters. The algorithm is simple and easy to program, so this paper chooses single neuron pid as the basic algorithm of digital governor. In the aspect of circuit design, the hardware circuit of this paper includes: single chip microcomputer minimum system, power supply module, speed signal shaping circuit, actuator driving circuit, voltage and current sampling circuit and other auxiliary circuits. Among them, the microprocessor is a 16-bit PIC24FJ64GA002 single-chip microcomputer, which uses its high-precision input capture module to collect the rotational speed signal, outputs the 16-bit resolution PWMM, and controls the switching on and off of the transistor in the actuator drive circuit. The current in the actuator is adjusted by changing the PWM duty cycle, and then the fuel supply of the diesel engine is adjusted. In software aspect, this paper introduces the basic idea of software design in detail, the core algorithm is single neuron PIDs. In the starting stage, a special sinusoidal oil quantity control curve is designed to realize the smoke control. In addition, the experiment on Cummins 220KW diesel generator set shows that the digital electronic governor based on single neuron pid has high sensitivity, high control precision and excellent speed regulation performance.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM314;TP273

【参考文献】