After several years collecting data, researchers could analyze the signals at each ocean location to determine the tidal characteristics. Altimeters on these missions acted as flying tide gauges. The Topex/Poseidon and Jason satellite altimeter missions were designed to observe and record this complexity. Waves are trapped and rotating around New Zealand, causing a high tide on one side of the islands with a simultaneous low tide on the other side. In the North Atlantic, we see waves mainly rotating anti-clockwise, with small amplitudes in the middle of the ocean and high amplitudes around the boundaries, especially along the coasts of northwest Europe and Britain. Even there, we see a complicated pattern as waves merge from the north and others separate northwards or southwards under Antarctic ice shelves. Waves run relatively unimpeded westward only around Antarctica. This animation shows the tides as a complex system of rotating and trapped waves with a mixture of frequencies. If our planet had no continents, tides would be hemispheric-sized bulges of water moving westward with the moon and sun. Two dimensional Vehicle simulation(vehicle kinematic,dynamic model,etc)Ī vehicle steering at a fixed angle and fixed speed,whose trajectory is exactly a circle.Ocean tides are not simple. One dimensional Car simulation(speed,distance,acceleration,friction,wind resistance,etc) PID twiddler(coefficient self optimization) Here's what have been done in this project: #PID Control and Vehicle Kinematic Simulation Here is another example of trajectories smooth algorithm: RRT Searching Using Vehicle Dynamic ModelĪfter calculating the shortest path,we want to smooth the path. !(Search/GraphSearch/doc/RTT%20and%20Smooth/search_test1-result-smooth-avoid obstacle.bmp) RRT Searching with Smooth Path and Obstacle Avoidance Using three dimensional(x,y,orientation) A* we can get the following path: In real life,situation is more complicated.For example,if we are heading to a place in rush hour and we need to go through a traffic light.Then perhaps we don't want to take a left turn because it's gonna wait for 5 minutes.Instead we may want to take a detour,which is faster then taking a left turn.This problem can be solved by adding a third dimension,which is orientation factor.Consider the following example: RRT Using Vehicle Dynamic Model,5D(x,y,theta,speed,steering speed) Notice that at the beginning of the red curve there’s a bump,which is caused by gyro drift and filtered out by Kalman filter later on.įollowing searching algorithms are implemented in this projects: Implemented a highly portable Kalman filter module in C.Here are some experiments of fusing gyroscope and accelerometer data.As we can see,Kalman filter does a good job fusing the sensors data and getting rid of gyro drifting.But it can not filter out the accelerometer spike noise. Showing all the text data with format 0 - 360 : speed /rpm | point1 point2 point3 point4 Showing how many point data are valid on status bar Options to filter out the bad point data Visualizing the dot data with selectable range from 1-10 meters Visualize the search trial route and shortest route after searchingĪ small program to debug the XV-11 LIDAR module.Built with Qt and it's cross platform on Windows/Mac/Linux. Use mouse to set start and target position arbitarily on the map Read bitmap directly as map,can easily create test map This is a GUI program for visualizing the SLAM algorithms,here are the current progress: Robotics Simultaneous Localization and Mapping Current Progress: - QSLAM ,Qt SLAM GUI Program - LIDAR Sensor Data Acquisition and Visulization - Kalman Filter - Map Search Algorithm Comparison(BFS, A* ,Dynamic Programming,RRT) - Smooth Algorithm - PID Simulator(one dimension) - Vehicle Dynamic Model Simulation
0 Comments
Leave a Reply. |