为改变煤炭生产过程中落后的排矸生产方式，推动煤炭生产工业的技术进步和实现节能减排，研究一种煤矸分拣机器人及其控制系统，提高井下煤矸石分选的自动化程度，不仅能够替代人工拣矸，解放劳动力，而且能够替代传统洗选方式，节约我国缺水地区水资源、减少环境污染，对促进煤炭行业的发展具有重要意义。本文以煤矸分拣机器人的桁架式机械臂本体为基础，对其控制系统进行了深入研究。 (1)确定煤矸分拣机器人的基本结构，依据机器人的工作原理和煤矸分拣工作的控制需求，提出煤矸分拣机器人的总体控制方案。 (2)研究模糊逻辑推理方法，根据识别区域待抓取矸石特性，构建待抓取优先级排序模型。为了解决拣矸胶带机上矸石的机械臂动态跟踪轨迹规划问题，采用比例导引法，建立机械臂拣矸过程的机械臂运动轨迹规划模型。考虑到大块矸石需要同步跟踪抓取的问题，基于七段式梯形曲线规划机械臂同步跟踪运动轨迹，为保证机械臂运动时间最短，基于遗传算法优化七段式梯形曲线中“匀加速”、“匀速”和“匀减速”阶段机械臂的运动时间，并给出轨迹规划整体流程。 (3)建立机器人单关节电机驱动滚珠丝杠系统的数学模型，并以此模型为被控对象，针对控制机械臂跟踪规划好的期望复杂轨迹问题，设计变论域模糊PID控制器，研究详细的设计过程，并在Matlab上对比传统PID控制器进行控制效果的仿真。 (4)研究煤矸分拣机器人的控制系统，搭建检测带速模块、控制模块和通信模块等硬件子系统，基于C#在Visual Studio开发环境中开发软件控制系统，采用以太网通信方式，确保上下位机的数据交互。在此基础上，以桁架式双机械臂为机器人本体，搭建煤矸分拣机器人控制系统平台，并进行试验，验证本文提出的控制方法的控制效果及开发的煤矸分拣机器人控制系统的正确性和可靠性。

In order to change the backward production mode of coal production in the coal production process, promote the technological progress of the coal production industry and realize energy conservation and emission reduction, study a coal shovel sorting robot and its control system to improve the automation degree of coal gangue sorting. It can replace manual picking, liberate labor, and can replace traditional washing methods, save water resources in China's water-scarce areas, reduce environmental pollution, and is of great significance to promote the development of the coal industry. Based on the truss type manipulator body of coal shovel sorting robot, this paper makes an in-depth study on its control system. (1)The basic structure of the coal shovel sorting robot is determined. According to the working principle of the robot and the control demand of the coal shovel sorting work, the overall control scheme of the coal gang sorting robot is proposed. (2)The fuzzy logic reasoning method is studied, and the priority ranking model to be captured is constructed according to the characteristics of the meteorstone to be captured in the recognition area. In order to solve the problem of dynamic tracking trajectory planning of the mechanical arm on the picking tape machine, the proportional guiding method is used to establish the mechanical arm trajectory planning model of the mechanical arm picking process. Considering the problem that large meteorites need to track and capture synchronously, the robotic arm synchronously tracks the motion trajectory based on the seven-segment trapezoidal curve. In order to ensure the minimum movement time of the manipulator, the "even acceleration" in the seven-segment trapezoidal curve is optimized based on the genetic algorithm. The movement time of the manipulator in the "constant speed" and "even deceleration" phases, and gives the overall process of the trajectory planning. (3)The mathematical model of the robotic single-joint motor-driven ball screw system is established. With this model as the controlled object, the variable-domain fuzzy PID controller is designed for the control robot to track the expected complex trajectory problem. The detailed design process is studied. And compare the traditional PID controller to control the simulation effect on Matlab. (4)Study the control system of coal gang sorting robot, build hardware subsystems such as detection speed module, control module and communication module, develop software control system based on C# in Visual Studio development environment, adopt Ethernet communication mode to ensure the upper and lower machine Data interaction. On this basis, the truss type double mechanical arm is used as the robot body, the control system platform of coal gang sorting robot is built, and the test is carried out to verify the control effect of the control method proposed in this paper and the correct control system of the coal shovel sorting robot developed. Sex and reliability.