A policy iteration method for improving robot assembly trajectory efficiency

Bolt assembly by robots is a vital and difficult task for replacing astronauts in extra-vehicular activities (EVA), but the trajectory efficiency still needs to be improved during the wrench insertion into hex hole of bolt. In this paper, a policy iteration method based on reinforcement learning (RL) is proposed, by which the problem of trajectory efficiency improvement is constructed as an issue of RL-based objective optimization. Firstly, the projection relation between raw data and state-action space is established, and then a policy iteration initialization method is designed based on the projection to provide the initialization policy for iteration. Policy iteration based on the protective policy is applied to continuously evaluating and optimizing the action-value function of all state-action pairs till the convergence is obtained. To verify the feasibility and effectiveness of the proposed method, a noncontact demonstration experiment with human supervision is performed. Experimental results show that the initialization policy and the generated policy can be obtained by the policy iteration method in a limited number of demonstrations. A comparison between the experiments with two different assembly tolerances shows that the convergent generated policy possesses higher trajectory efficiency than the conservative one. In addition, this method can ensure safety during the training process and improve utilization efficiency of demonstration data.     

Numerical study on combustion efficiency of aluminum particles in solid rocket motor

The combustion of aluminum particles in solid rocket motor plays an important role in energy release of propellants. However, due to the limited residence time, aluminum particles may not be burned completely, thus hindering the improvement of specific impulse. This study aims to explore the characteristics of aluminum combustion efficiency and its influencing factors by experiments and numerical simulations, providing a guideline for engine performance improvement. As an input of simulation, the initial agglomerate size was measured by a high pressure system. Meanwhile, the size distribution of the particles in plume was measured by ground firing test to validate the numerical model. Then, a two-phase flow model coupling combustion of micro aluminum particle was developed, by which the detailed effects of particle size, detaching position and nozzle convergent section structure on aluminum combustion efficiency were explored. The results suggest that the average combustion temperature in the chamber drops with increasing initial particle size, while the maximum temperature increases slightly. In the tested motors, the aluminum particle burns completely as its diameter is smaller than 50 μm, and beyond 50 μm the combustion efficiency decreases obviously with the increase of initial size. As the diameter approaches to 75 μm, the combustion efficiency becomes more sensitive to particle size. The combustion efficiency of aluminum particle escaping from end-burning surfaces is significantly higher than that from internal burning surface, where the particle combustion efficiency decreases during approaching the convergent section. Furthermore, the combustion efficiency decreases slightly with increasing nozzle convergent section angle. And theoretically it is feasible to improve combustion efficiency of aluminum particles by designing the convergent profile of nozzle.     

Prediction of electric-powered fixed-wing UAV endurance

To overcome the problems encountered in predicting the endurance of electricpowered fixed-wing unmanned aerial vehicles (UAVs),which were stemmed from the dynamic changes in electric power system efficiency and battery discharge characteristics under different operating conditions,the required battery power model and battery discharge model were studied.The required battery power model was determined using an approximate model of electric power system efficiency based on wind tunnel testing and the self-adaptive penalty function.Furthermore,current correction and ambient temperature correction terms were proposed for the trained Kriging model representing the discharge characteristics under standard operation,and then the discharged capacity-terminal voltage model was established.Through numerical integration of this model with the required battery power model,the electric-powered fixed-wing UAV endurance prediction model was obtained.Laboratory tests indicated that the proposed endurance model could precisely calculate the battery discharge time and accurately describe the battery discharge process.The similarity of the theoretical and flight test results reflected the accuracy of the proposed endurance model as well as the importance of considering dynamic changes in power system efficiency in endurance calculations.The proposed endurance model meeting precision requirements can be used in practical engineering applications.