基于DQN和K-means聚类算法的天然气站场仪表智能识别研究

Research on Intelligent Recognition of Meters in Natural Gas Station Based on DQN and K-means Clustering Algorithm

天然气站场中的仪表是工人和设备交互的窗口,可以反映工厂的运行状况;但是站场很多老式仪表不能远程读取示数,采用人工方法读取则浪费人力,需要对其进行智能化的读数研究;针对上述问题,采用了一种基于四足机器人作为载体运动控制,并通过深度强化学习(DQN)进行目标追踪任务和图像处理来读取仪表示数的新方法;首先通过改进的DQN算法的深度网络模型,根据仿真的环境中机器人学习效果,设计并调整动作奖励函数,设计机器人顶层决策控制系统;实现一维与二维状态参数输入下的仪表目标追踪任务;其次在仪表定位和仪表配准的基础上,通过K-means聚类二值化处理得到刻度分明的表盘;将图像进行内切圆处理,再在图像中间添加一根指针进行旋转,旋转过程中精确计算指针与表盘重合度最高的角度来得到对应刻度;经过实验表明,此算法可实现运动过程中仪表目标的精准追踪和降低计算时间,并大大提高了仪表追踪与识别的精度和效率,为天然气站场的仪表安全监控提供了有效保障。

Meters in a natural gas station is window for the interaction between workers and equipment,which can reflect the operation status of a plant.However,many old-fashioned instruments in the station yard cannot read the readings remotely,and manual reading is a waste of manpower,and it is necessary to carry out an intelligent reading research.Aiming at above problem,a new method based on quadruped robot is taken as the carrier motion control,the target tracking task and image processing by deep reinforcement learning(DQN)is adopted to read the instrument representation number.Firstly,through the deep network model of the improved DQN algorithm,the robot learning effect in the simulated environment is used to design and adjust the action reward function,and layout the top-level decision control system of the robot,the instrument target tracking under the input of one-dimensional or two-dimensional state parameters is realized.Secondly,on the basis of meter positioning and meter registration,the binarization of the K-means clustering is used to obtain a dial with clear scale;The image is inscribed circle,and then a pointer is added in the middle of the image to rotate,during the rotation process,the highest coincidence angle between pointer and dial is accurately calculated to obtain the corresponding scale.Experiments show that this algorithm can achieve the accurate tracking of instrument targets and reduce the calculation time during the movement process,and greatly improve the accuracy and efficiency of instrument tracking and identification,and it provides an effective guarantee for the instrument safety monitoring in natural gas stations.