During the simulation analysis of the discrete element method (DEM) for the alfalfa compression process, the input parameters in DEM software had a significant effect on simulation results. To obtain simulation parameters of the alfalfa with different moisture contents, a combination of angle of repose tests and simulation optimization design are presented in this paper. The repose angle of the alfalfa with moisture contents of 2.7%, 13.4%, 19.9%, 33.1%, and 74.5% was measured, and the results were 41.99˚, 38.30˚, 47.47˚, 56.31˚, and 63.09˚, respectively. Inclinometer tests, shear test, and restitution test were performed to evaluate the contact parameters. Taking contact parameters as the calibration object, the Plackett-Burman test was used to screen out which parameters had significant influence on the repose angle. The results of variance analysis showed the surface energy was the most significant parameter in the alfalfa repose angle test for each moisture content. Based on the Box-Behnken test, a second-order regression model of repose angle was obtained and the significance parameters were optimized and calibrated. The parameters calibrated in this paper can provide a reference for other simulations of alfalfa utilization.

No abstract

People always desire an embodied agent that can perform a task by understanding language instruction. Moreover, they also want to monitor and expect agents to understand commands the way they expected. But, how to build such an embodied agent is still unclear. Recently, people can explore this problem with the Vision-and-Language Interaction benchmark ALFRED, which requires an agent to perform complicated daily household tasks following natural language instructions in unseen scenes. In this paper, we propose LEBP -Language Expectation with Binding Policy Module to tackle the ALFRED. The LEBP contains a two-stream process: 1) it first conducts a language expectation module to generate an expectation describing how to perform tasks by understanding the language instruction. The expectation consists of a sequence of sub-steps for the task (e.g., Pick an apple). The expectation allows people to access and check the understanding results of instructions before the agent takes actual actions, in case the task might go wrong. 2) Then, it uses the binding policy module to bind sub-steps in expectation to actual actions to specific scenarios. Actual actions include navigation and object manipulation. Experimental results suggest our approach achieves comparable performance to currently published SOTA methods and can avoid large decay from seen scenarios to unseen scenarios.

No abstract

No abstract