Compared with the robots, humans can learn to perform various contact tasks in unstructured environments by modulating arm impedance characteristics. In this article, we consider endowing this compliant ability to the industrial robots to effectively learn to perform repetitive force-sensitive tasks. Current learning impedance control methods usually suffer from inefficiency. This paper establishes an efficient variable impedance control method. To improve the learning efficiency, we employ the probabilistic Gaussian process model as the transition dynamics of the system for internal simulation, permitting long-term inference and planning in a Bayesian manner. Then, the optimal impedance regulation strategy is searched using a model-based reinforcement learning algorithm. The effectiveness and efficiency of the proposed method are verified through force control tasks using a 6-DoFs Reinovo industrial manipulator.
The theoretical analysis of the charge exchange process in a spark ignition engine has been presented. This process has significant impact on the effectiveness of engine operation because it is related to the necessity of overcoming the flow resistance, followed by the necessity of doing a work, so-called the charge exchange work. The flow resistance caused by the throttling valve is especially high during the part load operation. The open Atkinson-Miller cycle has been assumed as a model of processes taking place in the engine. Using fully variable inlet valve timing the A-M cycle can be realized according to two systems: system with late inlet valve closing and system with early inlet valve closing. The systems have been analysed individually and comparatively with the open Seiliger-Sabathe cycle which is a theoretical cycle for the classical throttle governing of the engine load. Benefits resulting from application of the systems with independent inlet valve control have been assessed on the basis of the selected parameters: fuel dose, cycle work, charge exchange work and a cycle efficiency. The use of the analysed systems to governing of the SI engine load will enable to eliminate a throttling valve from the system inlet and reduce the charge exchange work, especially within the range of part load operation.
Effects of various friction stir processing (FSP) variables on the microstructural evolution and microhardness of the AZ31 magnesium alloy were investigated. The processing variables include rotational and travelling speed of the tool, kind of second phase (i.e., diamond, Al2O3, and ZrO2) and groove depth (i.e., volume fraction of second phase). Grain size, distribution of second phase particle, grain texture, and microhardness were analyzed as a function of the FSP process variables. The FSPed AZ31 composites fabricated with a high heat input condition showed the better dispersion of particle without macro defect. For all composite specimens, the grain size decreased and the microhardness increased regardless of the grooved depth compared with that of the FSPed AZ31 without strengthening particle, respectively. For the AZ31/diamond composite having a grain size of about 1 μm, microhardness (i.e., about 108 Hv) was about two times higher than that of the matrix alloy (i.e., about 52 Hv). The effect of second phase particle on retardation of grain growth and resulting hardness increase was discussed.