A5102786301- Soil Mechanics and Vehicle Dynamics
- Topic Modeling
- Natural Language Processing Techniques
- Reservoir Engineering and Simulation Methods
- Robotic Locomotion and Control
- Speech and dialogue systems
- CO2 Sequestration and Geologic Interactions
- Real-time simulation and control systems
- Groundwater flow and contamination studies
- Agriculture and Farm Safety
- Vehicle Dynamics and Control Systems
- Control and Dynamics of Mobile Robots
- Hydraulic and Pneumatic Systems
Beijing University of Civil Engineering and Architecture
2023
Stanford University
2017
Nanjing University of Aeronautics and Astronautics
2013
Abstract The operation of most engineered hydrogeological systems relies on simulating physical processes using numerical models with uncertain parameters and initial conditions. Predictions by such can be greatly improved Kalman‐filter techniques that sequentially assimilate monitoring data. Each assimilation constitutes a nonlinear optimization, which is solved linearizing an objective function about the model prediction applying linear correction to this prediction. However, if conditions...
In the past year, capabilities of Multimodal Large Language Models (MLLMs) have significantly improved across various aspects. However, constrained by quadratic growth computation in LLMs as number tokens increases, efficiency has become a bottleneck for further scaling MLLMs. Although recent efforts been made to prune visual or use more lightweight reduce computation, problem with increase still persists. To address this, we propose novel approach: instead reducing input LLMs, focus on...
Self-improvement in multimodal large language models (MLLMs) is crucial for enhancing their reliability and robustness. However, current methods often rely heavily on MLLMs themselves as judges, leading to high computational costs potential pitfalls like reward hacking model collapse. This paper introduces a novel, model-level judge-free self-improvement framework. Our approach employs controlled feedback mechanism while eliminating the need verification loop. We generate preference learning...
In order to solve the problems of complicated steering control unmanned vehicles in field and difficult on complex roads, we designed a wheel-track composite vehicle equipped with novel power differential mechanism dual driving, which drove through rotation rear two wheels. The was simple control, small size, able work under conditions such as hills, mountains, muddy land. Firstly, both speed force prevent from skidding into land stopping motionless. Secondly, kinematics model dynamics drive...