中文论文文摘

    大型工业锅炉对燃烧状态进行状态监控和燃烧量化评价，一直是燃烧领域关注的问题。由于燃烧本身的复杂性和测量手段的限制，使用常规的测量方法和诊断手段，到目前为止还没有形成一套完整意义上的燃烧过程状态监控和量化评价的方法。基于火焰图像处理为燃烧过程的状态监控开辟了一条新思路。
    本论文利用图像处理技术得到丰富的火焰信息，根据火焰特性提出了能反映火焰燃烧状况的火焰特征量，并提出了一种基于彩色CCD三基色比色方法来测量火焰温度场，构建了小波神经网络模型进行燃烧诊断，据此判断炉膛火焰燃烧情况并预测火焰燃烧趋势。利用DSP完成对图像的实时处理分析，使系统具有较高的实时性和可靠性。
    本文构架了基于TI公司的TMS320C6205 DSP芯片的实时图像采集处理系统。该系统能够克服传统基于PC机模式的图像处理系统的缺点，提高系统的实时性能。采用SAA7111和CPLD实现系统前端的图像采集以及图像的输出控制，用DSP实现图象处理算法，并通过PCI接口将运算结果传输给上位机，对锅炉火焰燃烧状况进行量化评价。文章详细介绍了系统的各个功能模块的具体设计方法以及系统软件的设计。
    最后，本文对本系统进行了总结，对算法的研究提出几点改进意见，并对系统的进一步研发提出了展望。

外文论文文摘

    It is paid attention to all the time that large rating power plants watch and control combustion situation and quantities the flame characteristic value in combustion field. Because of the complexity of combustion and restriction of measure means, so far it doesn’t form means of combustion process control and quantity yet up till now by routine measurement method and diagnostic means.
    This thesis analyzes flame traits and distills flame characteristic values, and demonstrates the feasibility that measure temperature using the flame image；Proposed a method of combustion diagnosis based on wavelet Neural Network, combined digital image processing technology and DSP, then the evaluation of flame burning situation and flameout forewarning function are implemented.
    This thesis structures an embedded system for flame image acquisition and processing based on TMS320C6205 DSP. This system can overcome the shortage of the traditional image processing system based on PC mode, and improve its performance in the field of real time. The output control of image is implemented by SAA7111 and CPLD. DSP is used to realize varied image processing arithmetic and interface with computer through the PCI bus. The dissertation analyzes the design technique of each function module in details and the implementing of system software is discussed.
    At the end of this paper, it makes a summary and provides some further advice to the arithmetic and this system.