Search

A new method is proposed to control bipedal robots to achieve flexible omni-directional motion and robust locomotion under complex disturbances, called the heuristics-based reinforcement learning (HBRL) framework. HBRL shows great training efficiency in simulation. Heuristic reference trajectories play a crucial role in HBRL, which guide the training process. Exploration rewards, leg-foot reset condition, and command curriculum are three significant components to optimize the training process. An estimator network is utilized to supply linear velocities and foot contact information. We train controllers on flat ground in simulation. To demonstrate robustness and versatility, the trained controllers were tested on BRAVER, a point-foot bipedal robot with three joints on each leg. The controllers enabled BRAVER to perform omni-directional locomotion with the maximum forward speed reaching 2 m/s. The robot could also maintain balance under external pushing and over uneven terrains. [ABSTRACT FROM AUTHOR]

This paper, based on a national key railway project crossing the high ground stress TBM construction section, studies the test results by means of the combination of refining three-dimensional numerical simulation and regression orthogonal design experiment and multiple linear regression method. The sensitivities of different supporting factors are studied, their advantages and disadvantages in controlling surrounding rock deformation are identified. At the same time, multiple linear regression equation of the experiment is obtained, which provides a fast calculation formula for structural optimization, so as to adjust the support parameters and optimizing the structure. In linear regression orthogonal experiment, the length of short anchor bolt, the length of long anchor bolt, the longitudinal spacing of steel arch, the circumferential and longitudinal spacing of anchor bolt, and the shotcrete grade and ground stress level are selected as the influencing factors. Regression orthogonal design experiment with L8(27) orthogonal table is conducted without considering interaction effect. The constant term and coefficients of partial regression term are calculated, and the multiple linear regression equation is established, and the results are analyzed in terms of variance and testing lose effectiveness of fit. The results show that the F value of ground stress is greater than F0.001, and the ground stress has a decisive influence on the deformation of the surrounding rock. The F values of the length of short anchor bolt, the longitudinal spacing of the anchor bolt and steel arch are all greater than F0.025, which have a significant impact on controlling the deformation of the surrounding rock. The established regression equation is significant, the test of misfit is not significant, so the regression equation has a good degree of fitting. Therefore, measures should be taken to release ground stress in design and construction first, and then the length of short anchor bolt, the spacing of steel arch and the longitudinal spacing of anchor bolt should be taken into account when selecting support parameters. [ABSTRACT FROM AUTHOR]

Stratum cavity collapse induced by water and sand leakage in a subway tunnel is a sudden urban geological disaster, which can cause serious casualties and economic losses. In response to the hazards of stratum cavity collapse induced by water and sand leakage around subway tunnels, a set of visual test devices with variable hole diameters were designed to simulate the stratum cavity collapse induced by water and sand leakage under comprehensive consideration of the size of the cavity diameter, clay thickness, clay strength, and so on. Based on pressure arch theory and unified strength theory, a critical calculation formula was established for stratum cavity collapse under two modes, of tensile shear and shear failure. Furthermore, the accuracy of the theoretical calculation formula was verified by experiment. The results show that the critical radius increases with the increase of clay strength and decreases with the increase of buried depth and friction angle. Clay soil is more prone to shear failure at the arch foot under the same conditions. The larger the soil sand mix ratio and the cavity diameter, the smaller the critical water pressure of soil collapse. The thicker the soil layer, the greater the critical water pressure. The theoretical analysis method has a clear physical meaning and the theoretical results agree well with the test results. The research results are of great significance for the warning of and prevention of such urban geological disasters. [ABSTRACT FROM AUTHOR]

When the tunnel passes through the slope area, once the slope stability changes or landslide disasters occur, large additional stress, deformation, or cracking are easily caused in the tunnel, which results in high risk to the tunnel operation. The deformation and failure mode of the tunnel induced by different types of tunnel-related landslides are different. However, there is no systematic classification of tunnel structure deformation and the failure mode in the tunnel-related landslides systems. By investigating the typical cases of tunnel diseases caused by landslide in China, this study analyzes the relationship between the tunnel-related landslide types and tunnel disease characteristics. It also summarizes five typical tunnel deformation failure modes and corresponding deformation characteristics and determines the relationship between tunnel-related landslide types and deformation failure modes. This study can provide a technical support for the identification of landslide types through the tunnel deformation characteristics in practical engineering applications. Finally, it provides a reference for tunnel location selection, the design of tunnel structure, and the treatment of landslide in landslide areas.

Chinese government puts forward a concept: five in one, putting environment in the same significant position as economy, politics, society and culture, which highlights the significance of environment. With the continuous increase of environmental awareness, many scholars pay more attention to green finance. However, the researches based on economic or social aspects, mostly stay on the theoretical level without studying how much people know about this concept. This paper, however, based on the questionnaire on Guangdong Undergraduates, aims to know about the deviation between undergraduates' awareness of the importance of green finance and the real situation. The results turn out that there are significant deviations in the following areas: concept, methods, reasons, purpose and representative principles, etc. Based on the results, the author tries to explain the reason of the existence of these deviations and the connection of enhancing education of environmental protection.

There are terminal devices with different processing rates and different types of communication networks in the Internet of Things. Most of the transmitted data have certain requirements for delay and bandwidth. How to ensure effective data transmission between communication devices in the Internet of Things, make the data flow on each link meet the requirements and prevent data overflow is of great significance to control the flow in the Internet of Things. In this paper, the traffic scheduling with multiple QoS constraints is converted as a multiple objective optimization. The mathematical model of multiple objective optimization is solved by a multiple objective particle swarm optimization algorithm to obtain Pareto optimal solution. The experimental results on Mininet show that the proposed traffic scheduling is superior to previous ones. [ABSTRACT FROM AUTHOR]