Error Analysis And Algorithm Research Of Multi-USV Collaborative Navigation

As one of important tools of safeguarding the country’s military security and territorialsovereignty, USV is being more and more focused. Since the difficulty of tasks is gettingharder, low-cost, high-performance multi-USV system now is a hot issue. High-precisionnavigation and positioning is a key point that the multi-USV system could complete tasks. Inthis paper, the collaborative navigation and positioning technique of multi-USV system isstudied. The main contents are as follows.1. Research on the SINS/DVL integrated navigation system of the leader USV in themaster-slave collaborative navigation system. The basic working principle of SINS and DVLare analyzed, the basic INS equations and speed calculating formula are deduced. And theerror model of DVL is established. SINS error model of Euler platform error angle underlarge misalignment angle is deduced. Then SINS/DVL integrated navigation model is set up,and simulation is carried out.2. The factors of influencing the collaborative navigation location accuracy, includingsound velocity error, clock synchronization and time delay, is analyzed quantitatively or/andqualitatively by studying their effect on the ranging accuracy. And the error analysis of deadreckoning method, which is used in the follower USV of the collaborative navigation system,is made.3. There are two collaborative location methods, multi-leader and single-leader.Algorithm of these two methods is researched. The positioning principle is introduced,positioning model established and collaborative navigation EKF algorithm presented.Simulation result shows that the algorithm could effectively locate the follower. Besides, theobservability of single-leader collaborative navigation system is analyzed by using nonlinearobservability theory. The analysis proves that path planning must be taken to increase thelocation accuracy when using EKF.4. Since in practice there is time delay that affects the location accuracy of themulti-USV system, the collaborative navigation algorithm based on time delay is presented.Firstly, predictive error based on state compensation method is deduced, and the DEKFalgorithm is presented. Simulations show that the algorithm can compensate the collaborativelocation error effectively. Since the estimation of state compensation method is not optimal,another method, measurement update, is introduced. The time delay is taken into the stateequation and measurement equation, and the system equation is reconstructed. Then state estimation equation, filter gain and state estimation mean-square deviation are deduced.Simulation is made and the results show that the measurement update method is better thanstate compensation method.