本课题研究面向3C行业中柔性电路板的机器人装配演示编程技术，旨在通过拖动机器人装配演示，实现自动编程使得机器人能适应装配部件差异而根据装配状态智能地选择动作以完成类人的装配工作，从而将人的技能快速地转移给机器人。具体的研究内容与技术方案包括：测量演示过程中的力和位移信息并分析其融合关系；研究基于力位信息的装配状态识别方法，建立装配过程的隐形马尔可夫模型，由演示数据辨识模型参数，为机器人智能选择装配动作提供依据；设计转换装配状态的基本操作规划方法，使用高斯混合模型和高斯回归方法提取演示动作特征，进而辨识基本操作参数，实现机器人力位融合的基本操作规划；研究机器人力位融合运动控制方法，设计基本操作平滑方法，开发演示编程软件；搭建试验平台，利用试验结果优化演示编程技术。课题的研究成果将实现机器人装配的简单、快速自动编程，无需用户具备编程专业知识，为3C行业装配线的机器人换人提供技术支持。

This project focuses on the programming by demonstration (PBD) for flexible printed circuit (FPC) assembly in 3C industry. It aims to endow the robot with the ability to assemble FPC as human does with little effort in programming. Thus the skill of the human can be transferred to the robot easily. The research consists the following contents. First, the displacement and force are measured during the assembly performed by human guiding the robot. And the fusion between the force and displacement on the assembly are analyzed. Second, the key assembly states are identified based on the force and displacement. Correspondingly, the assembly process is modeled via Hidden Markov Model (HMM). This model guides the robot to choose proper movements after its parameters being identified according to the demonstrations. Third, the dynamic movement primitives together with Gaussian Mixture Model (GMM) and Gaussian Mixture Regression (GMR) are utilized to generalize the basic operations between the two key states. Forth, the control algorithms for the motion with the force and displacement fusion are designed. Then the software combining the assembly process model, the basic operations and the smoothness between operations is developed. At last, the experiments are carried out to verify and optimize the proposed PBD. The achievement of this project will provide a fast and simple programming method for the robot assembly without the requirements of professional programming skill, and thus support that robot substitutes human in 3C industry.