No abstract

When robots are built with state-driven motors, task-planning increases in complexity and difficulty. This type of actuator is difficult to control, because each type of control position/force requires different motor parameters. To solve this problem, we propose a state machine-driven hybrid position/force control architecture (SmHPFC). To achieve this, we take the classic hybrid position/force control method, while using only PID regulators, and add a state machine on top of it. In this way, the regulators will not help the control architecture, but the architecture will help the entire control system. The architecture acts both as a parameter update process and as a switching mechanism for the joints’ decision S-matrix. The obtained control architecture was then applied to a 5DOF serial manipulator built with Festo motors. Using SmHPFC, the robot was then able to operate with position or force control depending on its designated task. Without the proposed architecture, the robot joint parameters would have to be updated using a more rigid approach; each time a new task begins with new parameters, the control type would have to be changed. Using the SmHPFC, the robot reference generation and task complexity is reduced to a much simpler one.

This paper presents a hybrid force/position control. We developed it for a hexapod walking robot that combines multiple bipedal robots to increase its load. The control method integrated Extenics theory with neutrosophic logic to obtain a two-stage decision-making algorithm. The first stage was an offline qualitative decision-applying Extenics theory, and the second was a real-time decision process using neutrosophic logic and DSmT theory. The two-stage algorithm separated the control phases into a kinematic control method that used a PID regulator and a dynamic control method developed with the help of sliding mode control (SMC). By integrating both control methods separated by a dynamic switching algorithm, we obtained a hybrid force/position control that took advantage of both kinematic and dynamic control properties to drive a mobile walking robot. The experimental and predicted results were in good agreement. They indicated that the proposed hybrid control is efficient in using the two-stage decision algorithm to drive the hexapod robot motors using kinematic and dynamic control methods. The experiment presents the robot’s foot positioning error while walking. The results show how the switching method alters the system precision during the pendulum phase compared to the weight support phase, which can better compensate for the robot’s dynamic parameters. The proposed switching algorithm directly influences the overall control precision, while we aimed to obtain a fast switch with a lower impact on the control parameters. The results show the error on all axes and break it down into walking stages to better understand the control behavior and precision.

No abstract

The paper presents a VIPRO versatile, intelligent and mobile latform for robots developed through adaptive networked control, using an original virtual projection method which involves the representation of modern mobile robots in a 3D virtual environment. The Cartesian robot workspace, which is learned using an adaptive neuro-fuzzy control for the prediction of desired references was generate. In correlation was used a strong robotic simulator, an open architecture system and adaptive networks over the classical control system of the robot. It was developed an intelligent control interfaces that apply advanced control technologies adapted to the robot environment such as control by artificial intelligence techniques using adaptive neuro-fuzzy inference systems, extended control (extenics), neutrosophic control, human adaptive mechatronics. The results lead to a high level real-time simulation platform as a new component, alongside the existing ones on the IT market, for modelling the interactions the robots in hazardous or challenging environments. Index Terms-Intelligent control interfaces, virtual projection method, adaptive networked control, tracking trajectory control, contradictory problems, extenics control.