摘要
为实现在实验室内快速定位气体泄漏源,提出了一种基于传感网络和PC_DSR GA的泄漏源定位系统。泄漏源定位是指根据气体扩散特点结合标记点坐标及该点气体浓度差异反算泄漏源,为此设计了基于CSMA/CD及自适应CRC校验退避的485总线通信协议,可更及时准确地获取标记点气体浓度,配备云台摄像头通过特征识别技术计算标记点坐标,提出一种基于高斯模型和PC_DSR GA的气体泄漏源反算算法。该算法以破坏自修复机制抑制GA的早熟,以保优策略保证种群最终收敛,实现更精确的定位泄漏源。实验结果表明,该系统传感网络可靠性强;所提算法与传统GA、混合遗传-NelderMead算法相比泄漏源定位精度更高,平均相对误差低于5%。
In order to realize the rapid location of gas leak sources in laboratory,a leak source location system based on sensor network and population conservation damage self-repair(PC_DSR)genetic algorithm was proposed.Leakage source localization refers to the inverse calculation of the leakage source according to the gas diffusion characteristics combined with the coordinates of the marker point and the gas concentration difference at that point.For this reason,the system designs a 485 bus communication protocol based on CSMA/CD and adaptive CRC check backoff,so it can obtain the gas concentration of the marker point in a timely and accurate manner.The system is equipped with a PTZ camera to calculate the coordinates of the marker points by feature recognition technology,and a gas leakage source inverse algorithm based on the Gaussian model and the PC_DSR genetic algorithm is proposed.The algorithm suppresses the prematureness of the genetic algorithm by destroying the self-repair mechanism,the guarantee strategy ensures the final convergence of the population,and achieves a more accurate positioning of the leak source.The experimental results show that the sensing network of the system is highly reliable.Compared with the traditional genetic algorithm and the hybrid genetic-NelderMead algorithm,the proposed algorithm has higher positioning accuracy and the average relative error is less than 5%.
作者
黄晓明
王忠华
HUANG Xiaoming;WANG Zhonghua(College of Information Engineering,Nanchang Hangkong University,Nanchang 330063,China)
出处
《实验室研究与探索》
CAS
北大核心
2020年第7期128-132,157,共6页
Research and Exploration In Laboratory
基金
国家自然科学基金(61362036)
研究生创新专项基金(YC2018025)
创新创业教育课程培育项目(KCPY1620)。
