随着高精度的无线室内定位需求的日渐增长，人们对定位的精度和实时性提出了更高的要求。目前超宽带（UWB）技术具有传输速率高、系统容量大和抗多径能力强等优点，但在复杂室内环境下测距误差波动大，定位精度受多径和非视距效应影响严重。所以，研究室内非视距环境下的定位算法，提高定位精度具有非常重要的意义。 为了减小复杂环境下非视距误差的影响，提高室内定位的精度，本文研究了UWB定位系统的基本原理，分析了UWB信道模型。基于异步TOA算法，提出一种改进的SDS-TWR测距算法，在信号传输基础上增加一组传输信号，通过使用两次重叠TWR方法实现初步测距。基于CHAN和Taylor协同算法，对测距初值进行计算，采用加权系数优化协同算法的计算结果，得到较精确的位置坐标。通过改进的UKF算法，对协同定位的坐标值进行滤波处理，在比例修正采样策略的基础上对UKF算法的状态更新方程进行修正，从而减小每次的迭代误差和采样的非局部效应，避免NLOS误差的累积。对改进的SDS-TWR测距算法，协同定位算法以及改进的UKF算法进行仿真分析。仿真结果表明，改进的测距算法由于时钟漂移引起的测距误差减小，测距值更接近于真实值；协同定位算法收敛性能优于单一算法，测距值标准误差相同的情况下具有更小的定位误差；改进的UKF算法收敛精度高，误差小于传统的滤波算法。 基于DWM1000模块搭建超宽带室内定位系统，测试并分析其定位精度。测试结果表明，系统对静止目标定位误差小于12cm，对运动目标定位误差小于35cm。改进的定位算法具有较高的定位精度，可以为UWB室内定位提供一定理论参考。

As the demand for the high-accuracy indoor location increases, higher requirements for the accuracy and real-time performance of location have been put forward. Ultra-wideband (UWB) technology has many advantages, such as high transmission rate, large system capacity and strong anti-multipath ability. However, the ranging value has a wide error range in the complex indoor environment, and positioning accuracy is seriously affected by multipath and non-line-of-sight effects. Therefore, it is of great importance to study the location algorithm in indoor non-line-of-sight environment and improve the location accuracy. In order to reduce the influence of NLOS error in the complex environment and improve the accuracy of indoor location, the basic principles of UWB positioning system are studied with UWB channel model analyzed. Based on the asynchronous TOA algorithm, an improved SDS-TWR ranging algorithm is proposed, which adds a group of transmission signals on the basis of the signal transmission and achieves the initial ranging value by using two overlapping TWR methods. On the basis of the CHAN and Taylor cooperative algorithm, the initial ranging value is calculated, and the weighted coefficients are applied to optimize the calculation results of the cooperative algorithm to obtain more accurate position coordinates. An improved UKF algorithm is proposed, which filters coordinate values of cooperative positioning. On the basis of proportional modified sampling strategy, the state update equation of UKF algorithm is modified to reduce the iterative error of each time and the non-local effect of sampling and avoid the accumulation of NLOS error. The improved SDS-TWR ranging algorithm, cooperative localization algorithm and improved UKF algorithm are simulated and analyzed. The simulation results show that the improved ranging algorithm reduces the ranging error caused by clock drift, and the ranging value is closer to the real value. The convergence property of the cooperative location algorithm is better than that of the single algorithm, and the location error is smaller with the same standard error of the ranging value. And the improved UKF algorithm has high convergence accuracy with the error smaller than that of the traditional filtering algorithm, which verifies the validity of the improved algorithm. Based on the DWM1000 module, the positioning system is built, and its positioning accuracy is tested and analyzed. The results of experiments show that the positioning error of the system is less than 12 cm for the stationary target and 35 cm for the moving target. The improved location algorithm has high positioning accuracy, which can provide certain theoretical reference for UWB indoor positioning.