This paper presents a novel algorithm for simultaneous position and interaction force control. In the classical algorithms, position and force control are executed concurrently by switching between two separate controllers: the position and force controller. Thus, one can consider the control system working in two modes, namely the position control and force control modes. Switching between these two modes often leads to oscillations in the controlled position and force. Therefore, the safe interaction between a controlled mechanical system and its environment is jeopardized. The above issues are tackled in this study by introducing a new control strategy. The proposed algorithm combines position and force control into a single controller, in which the transition between position and force control is smooth, removing the oscillations of classical methods. Therefore, the safe interaction between a mechanical system and its environment is enabled. In addition, using this method one can equip actuators with a control system capable of performing both position and force control. Thus, a step towards “smart actuators” is possible.

In shared human-robot environments, control systems operate based on the information about both human and robot activities to facilitate the successful collaboration between the two. This paper contributes to the emerging field of human-robot collaboration (HRC) by unifying human action recognition (HAR) and high-level robot control technique into single control system. Approach in this paper includes artificial neural network based classifier for recognition of human activity and task-based control as an example of high-level control technique. Classifier is developed based on the data from wearable sensors attached on the human arms. Recognized human activity is used as the input for the selection of functions that describe robot's activity (task). This papers combines both the theoretical approach to the task-based control and it's synergy with HAR while the developed artificial neural network classifier is experimentally validated.

SUMMARY This paper proposes a bilateral control structure with a realization of the force derivative in the control loop. Due to the inherent noisy nature of the force signal, most teleoperation schemes can make use of only a proportional (P) control structure in the force channel of the bilateral controllers. In the proposed scheme, an α–β–γ filter is designed to smoothly differentiate the force signal obtained from a reaction force observer integrated to both of the master and slave plants. The differentiated force signal is then used in a proportional-derivative (PD) force controller working together with a disturbance observer. In order to design the overall bilateral controller, an environment model based on pure spring structure is assumed. The controller is designed to enforce an exponentially decaying tracking error for both position and force signals. With the presented controller design approach, one can independently tune the controller gains of the force and the position control channels. The proposed approach is experimentally tested in a platform consisted of direct drive linear motors. As illustrated by the experiment results, the contribution of the PD control in the force channel improves the teleoperation performance especially under hard-contact motion scenarios by attenuating the oscillations, hence, improving the transparency when compared to the structures using only a P force control.

We generalize results concerning averaged controllability on fractional type equations: system of fractional ODEs and the fractional diffusion equation. The proofs are accomplished by introducing appropriate Banach space in which we prove observability inequalities.

The main goal of this study is comparative analysis of different methods used in design of digital fractional-order differentiator and integrator. The fractional-order digital differentiator or integrator can be described (in continuous time domain) with a transfer function H(s)=s^ɑ, where ɑ is a real number. To implement digital differentiators and integrators of arbitrary order the main step is the discretization. There are two common approaches of discretization. In this paper the direct and indirect discretization are presented but the emphasis will be on the indirect method, where the generating functions can be obtained through bilinear transformation, Al-Alaoui operator, Euler's backward operator and stable Simspon operator. The main differences between alternatives will be provided through analysis and comparison of their frequency responses - magnitude-frequency response, phase-frequency response and Nyquist diagrams.

One of the main requirements for motion control systems is their robustness. There are several previously introduced control methods possessing the aforementioned requirement. The algorithms designed in acceleration control framework based on disturbance observer (DOB) belong to that group. The DOB introduced in eighties by Ohnishi and his group has been discussed considerably in the literature and applied in numerous applications. However, in majority of them, the control algorithms have been designed in continuous time domain, even though today they are almost always implemented on digital platforms. In addition, a discussion about discrete implementation of the designed algorithms, and influence of the parameters of discrete systems on the control performance, appeared only in several previous works. This paper aims to fill part of that gap, by discussing discrete implementation of the control algorithms which are well developed in continuous domain. The given results will provide an opportunity to give recommendations for engineers employing DOB-based control in practice.

This paper describes an Internet Of Things-based wearable system for physical rehabilitation monitoring and characterization. The system is recording movement data with 3-axial accelerometer and gyroscope sensors. Data recorded by the sensors are used for characterization of movement, thus allowing for monitoring and estimation of the patients' state at all times. Main three parts of the system are: data acquisition unit, data processing unit, and cloud-based service for remote access to data. Hardware implementation is described and shown for each of the three parts. The system is demonstrated for monitoring of elbow rehabilitation. Results show that the device can be used for highly precise and accurate monitoring of elbow flexion and extension characteristics, thus allowing for remote rehabilitation tracking through the use of the cloud-based service.