针对自动驾驶、智能港口和智慧城市等新兴应用在复杂环境下鲁棒高精度定位需求，将全球卫星导航系统（GNSS）、惯性导航系统（INS）、激光雷达（LiDAR）等互补定位技术进行组合是目前公认的最佳方案。作为唯一具有全球统一时空框架的定位技术，GNSS在组合系统中起着主导性的作用；但在复杂环境中，实时动态差分定位（RTK）和精密单点定位（PPP）两种GNSS高精度定位技术性能显著恶化；而现有GNSS/LiDAR/INS组合系统多为松耦合架构，无法改善复杂环境下GNSS高精度定位性能，因而限制了组合系统性能的进一步提升。为此，论文以提高复杂环境下GNSS/LiDAR/INS组合系统中GNSS高精度定位性能为目标，对基于LiDAR/INS辅助的RTK和PPP算法展开了研究，主要工作包括： 首先，针对由复杂环境下GNSS观测量质量恶化导致的传统模糊度域搜索RTK算法固定率与定位精度下降的问题，提出了一种基于粒子滤波与均方残差（MSR）目标函数的位置域搜索RTK算法。该方法通过粒子滤波对基于载波相位观测量的位置域后验概率分布进行逼近，因而对周跳和上下星不敏感，且未使用码伪距观测量因而具有较强抗多径能力，可显著提高复杂环境下RTK固定率与定位精度。 其次，针对城市峡谷等复杂环境中可见卫星数少、分布几何差等造成的RTK性能恶化问题，提出了一种利用随路建图LiDAR特征作为伪卫星的LiDAR/INS辅助RTK算法。该算法利用GNSS/LiDAR/INS的互补性，在由RTK为当前LiDAR特征建立全局坐标的同时将已建图的LiDAR特征观测与RTK紧耦合。理论分析表明该算法可加强复杂环境下RTK几何约束，提高其固定率与定位稳定性；进而，设计了一种模糊度域/位置域搜索平行滤波架构对该辅助RTK算法进行实现。 然后，针对复杂环境下另一种GNSS高精度定位技术PPP性能提升的需求，充分考虑PPP收敛慢、分米级精度的技术特点，提出了一种利用LiDAR帧间相对位姿变化作为观测约束的LiDAR/INS辅助PPP算法。该算法通过PPP/LiDAR/INS的紧耦合加强了复杂环境下PPP几何约束，可有效抑制卫星数不足时PPP定位的发散，提高其在复杂动态环境下的收敛稳定性。 最后，基于自研仿真平台与多传感器硬件平台对各算法性能进行了测试评估，验证了上述算法在提升复杂环境下GNSS高精度定位精度和可用性方面的有效性。

Currently, many emerging positioning applications such as autonomous driving, unmanned ports and smart cities pose an urgent demand for robust high-precision positioning technology in GNSS-difficult environments, and the integration of the multiple complementary positioning technologies such as GNSS, INS and LiDAR becomes a consensus best solution to meet such demand. As the only technology that can provide positioning results in global frame, GNSS high-precision positioning technology plays a dominant role in the integrated system. However, the performances of two GNSS high-precision positioning technologies, namely RTK and PPP, degrade greatly in GNSS-difficult areas, and the current multi-sensor integration positioning systems are mainly implemented in a loose coupling mode which does nothing to the improvement of the GNSS high-precision positioning in GNSS-difficult areas. Therefore, aiming at improving the performance of GNSS high-precision positioning based on the GNSS/LiDAR/INS integration positioning system in GNSS-difficult areas, researches on LiDAR/INS-aided GNSS high-precision positioning technologies have been carried out in this thesis. The main research work of the thesis can be listed as follows: Firstly, to settle the problem of RTK fix rate and positioning accuracy degradation caused by GNSS measurement quality deterioration in GNSS-difficult areas, a particle filter-based MSR RTK algorithm in position domain is proposed. The algorithm adopts particle filter to approximate the posterior probability density of position space based on carrier phase measurements, and thus, is insensitive to cycle slips and changes of visible satellites. Besides, code pseudorange measurements are not used in the calculation, and thus, the algorithm has the resistance to multipath. The proposed method can improve RTK fix rate and positioning accuracy in GNSS-difficult areas. Then, to solve the problem of RTK performance deterioration caused by the challenges such as low satellite availability and poor satellite geometric distribution in GNSS-difficult areas, a LiDAR/INS-aided RTK algorithm using mapping LiDAR features as pseudo satellites is proposed. The algorithm takes advantages of the complementarity among GNSS, LiDAR and INS, and utilizes RTK to register LiDAR features while the measurements of the registered LiDAR features are integrated to RTK. The theoretical analyses prove that the algorithm can enhance RTK geometric constraint, and thus, improve RTK fix rate and stability in GNSS-difficult areas. A parallel-filter-based scheme in ambiguity-position-joint domain is further proposed to implement the LiDAR/INS-aided RTK algorithm. Next, to improve the performance of another GNSS high-precision positioning technology, namely PPP, in GNSS-difficult areas, we take the PPP characteristics of slow convergence and positioning precision at decimeter level into full consideration, and address a LiDAR/INS-aided PPP algorithm based on the constraint of LiDAR relative pose variations derived by frame matching. The algorithm enhances the geometric constraint of PPP with the use of the PPP/LiDAR/INS tight integration, and thus, can suppress the positioning divergence in the situation with insufficient visible satellites and improve the stability of PPP convergence in GNSS-difficult areas. Finally, the experiments based on the self-developed simulation platform and multi-sensor hardware platform have been conducted to evaluate the performances of the proposed algorithms, and validated the effectiveness on the improvement of GNSS high-precision positioning accuracy and availability in GNSS-difficult areas.