Quadruped walking robots have been a major focus of many research groups in recent decades. Walking robots offer many advantages over their traditional wheeled counterparts, primarily in the navigation of uneven, unsteady, or otherwise unusual terrain (de Santos 2006). Among walking robots, quadruped robots offer the minimum number of legs necessary to provide a stable support platform at all times during locomotion by moving only one leg at a time while all other legs remain in contact with the ground. As a result, quadruped robots can serve as stable and reliable platforms for a number of applications, ranging from surveillance to assisting the elderly. Most applications which can make use of quadruped walking robots can also greatly benefit from a robot capable of moving in any direction without the loss of speed, stability, or mobility.
R. B. McGhee and A. A. Frank developed a measure of stability called the static stability margin based on the relative locations of a robots center of mass and the positions of its points of contact with the ground (McGhee and Frank 1968).Chang-de Zhang and Shin-Min Song conducted substantial research into the determination of continuous stable walking gaits using the stability method developed by McGhee and Frank (Zhang and Song 1989). Numerous stability measurements have been developed in recent years, taking into account different relevant aspects of robotic motion. A case study comparing the validity and relative usefulness of various stability methods was conducted by Pablo Gonzalez de Santos who concluded that the Normalized Dynamic Energy Stability Margin (NDESM) is the most accurate stability margin of all the tested stability margins for every set of control conditions tested (de Santos 2006).
- Initial Design and Control of an Omni-Directional Quadruped Robot
One of end objectives for the spider robot is to be an agent capable of walking from a given start position to a goal position. The complexity of this portion of the project stems from the fact that the agent has no foregoing information of the terrain between the start and end positions. This terrain must be analyzed as the agent moves. In order to do this we will design a system that will allow the agent to understand its environment and make intelligent decisions regarding which direction to move. Weak AI in the form of behavioral based artificial intelligence which will allow the agent to make mistakes and make intelligent decisions in the place of seemingly random ones. This will be implemented using a subsumption architecture which will decompose the complicated intelligent behavior into many simpler modules which are organized into layers. Each layer will implement a particular goal of the agent, and higher layers are increasingly abstract. Each layer’s goal subsumes that of the underlying layers. Feedback will be given to the agent through the environment. The machine perception will come from IR sensors mounted on the top of the agent used to deduce the obstacles in the robots path.
Masters Student Scott Prince
Masters Student Kara West