Mobile robot fixed position is affirmatory its are located in foregone environment positional process, it is implementation shift robot the key of automatic navigation ability. What use sensor type according to robot place is different, its locate means differs somewhat. The sensor with at present wider application has camera of speedometer, ultrasonic, laser, camera, infrared ray, deepness, GPS to locate the system is waited a moment. With its corresponding robot locates the technology can divide two kinds big into technology of absolutely fixed position, relative fixed position. So called absolutely fixed position is to point to use navigation mark, advocate (by) move label, map to match, the technology such as GPS has fixed position, precision is taller. And relative fixed position is to point to conclude through magnanimity robot relative to the direction at position of rest and distance the positional information with a current robot, call boat again computative law. 1. The problem of robot fixed position that is based on the fixed position technique that the map matchs to be based on a map to match basically emphasizes particularly on of the position with analytic may located on the map robot search for and discern, its depend on a robot can feeling the photograph of positional environment information in the positional information that obtains located and local environment and foregone map to match mainly. In addition, the robot that is based on a map to match locates normally need and photograph of method of other fixed position are united in wedlock to realize fixed position then. Be based on deepness vision to describe a robot located geometrical map of the environment and be based on Ka Erman filter] , the scanning of particle filter wave matchs location method is conclude with probability the classical delegate that methodological photograph combines, and the application that has a success. 2. The guide of fixed position technology that is based on guide to mark] specific it is to point to those who have striking feature, and a general designation of a kind of object that can be distinguished by sensor of installation of the place on the robot. The guide of factitious set has the situation that him itself secures in the three dimensional space with located robot. Accordingly, the core task that the robot locates wants reliable ground, apace to differentiate namely outlet mark, calculate a robot is located the real situation in the map. The discretion of precision of real fixed position basically depends on what mark to guide is accurate differentiate and the accurate and rapid rate that draws to environmental place information. 3. Be based on probability estimated boat technology of computative fixed position is in process of robot fixed position, put in a lot of not certain factors. The noise interference of the accumulation that for instance robot itself has error of uncertainty, speedometer, sensor and robot are located environmental complexity, sealed sex is waited a moment. Anyhow, because these do not determine the presence of the element, make robot fixed position becomes relatively more complex. Some closer year, a lot of research scholars go to probability theory application among robot fixed position. The data that nuclear idea thinks even if collects till the foundation is current is foregone condition, next recursive estimation condition space is posterior probability density. Among them, the probability estimation that is based on particle filter wave realizes robot fixed position to have applied perspective more. Particle filter wave, also call alignment Monte Carlo, it is 20 discipline a kind of when development of the later period in 90 time rises brand-new filter wave is algorithmic, its nucleus idea thinks even if state probability distribution originally with random sample. The probability model photograph that is the same as algorithm of particle filter wave robot motion, perception joins the person such as Dallert, advanced the idea that robot Monte Carlo locates. Nuclear idea thinks even if estimate the place with the located likelihood of the robot with a group of filter, be in the probability of this position namely. Each filter correspondence has a place, recycle observation adds authority processing to every filter, make then the most possible the probability of located position is higher and higher. 4. Conclusion and look into up to now, the fixed position problem of indoor robot still is one of core problems that current robot learns to research a field, now the working key after is the fixed position precision that raises a robot and robot complete implementation locates independently, and be perfected even improve the stability that applies at fixed position algorithm, fast sex and effectiveness. The fixed position that robot fixed position uses SLAM method to carry out the real time compose that pursues to sealed environment ground to build application to match at the map, and the heavy sampling with more efficient research and the particle filter wave that get used to sampling oneself locate algorithm has better applied prospect.