(1) Focusing on indoor robot positioning, introduce and analyze common positioning technologies and characteristics. The principle, model and evaluation index of UWBtechnology are briefly introduced, which provides a theoretical basis for the follow-up research content.(2) In terms of ranging in indoor wireless environment, a UWB-based asynchronous no reference is proposed to solve the problems of complicated TDOA ranging, low stability and low accuracy of UWB positioning. A new method of ranging. The method firstly optimizes the UWB communication scheme based on TDOA, and introduces the actual base station distance as a correction factor to build a ranging algorithm for the base station without reference nodes under the asynchronous clock, and directly derives the arrival distance difference through the base station ranging information, reducing the complexity of calculation. Then, a dynamic filtering method is designed for the abnormality and accuracy of the ranging value, and a bilateral filtering algorithm is proposed to double-weight the proximity and relevance of the TDOA ranging value to further improve the ranging accuracy. The experimental results show that the ranging accuracy of this method is concentrated within 10cm, which can meet the ranging requirements of the actual system on the premise of ensuring real-time performance and stability.(3) In the aspect of indoor positioning, in view of the high complexity and low positioning accuracy of UWB positioning fusion algorithm, an improved Chan positioning algorithm based on asynchronous no reference ranging is proposed. Firstly, the asynchronous no reference ranging values are initially screened and Chan localization is realized, and the outliers are screened for the ranging values. Then, a bilateral filtering algorithm is proposed to weighted the coordinates calculated by the Chan localization algorithm, and the residual discriminant formula is used. Reliability identification of the current positioning result. The final positioning results show that: when the positioning error is 5cm, the cumulative distribution of the Chan algorithm is 0.52, and the improved cumulative distribution function in this thesis can reach 0.92. Under the premise of ensuring the computational complexity, the positioning accuracy is significantly improved.(4) Finally, a UWB-based indoor robot positioning system is designed and developed.In this thesis, an indoor robot positioning platform with four base stations and one label isdesigned, and the asynchronous no reference ranging algorithm is implemented in the underlying driver code based on STM32 and BU01 modules. The upper computer softwareplatform implements the positioning algorithm and designs the display interface.Experiments show that the positioning system designed in this thesis is easy to operate,stable, and has small errors, which can meet the actual positioning needs.