Monte Carlo simulation method for behavior analysis of an autonomous underwater vehicle

Abstract

The paper presents the application of Monte Carlo simulation in the behavior analysis of an autonomous underwater vehicle. Due to the highly nonlinear dynamics and existence of uncertain parameters in the models, there is not a straightforward method to analyze the behavior of an autonomous underwater vehicle. The objective of this article is to introduce a Monte Carlo campaign for an autonomous underwater vehicle 6-degree-of-freedom model to examine the effects of uncertain parameters on the mission objectives. Uncertainties in the model are considered in several categories, consisting of hydrodynamic and added mass coefficients, control instruments (sensors and actuators), environmental conditions and initial conditions. Monte Carlo simulations are run for a typical autonomous underwater vehicle moving from the water surface to reach a predetermined depth and heading during the mission time. For this purpose, 6-degree-of-freedom software is developed in C++ which is a fast and visual programming software. Using an example, it is shown that simulation results can be used for tuning of guidance algorithm. Moreover, the proposed concept is applicable for analysis of other types of autonomous ocean systems.

Real‐time harmonic estimation using a novel hybrid technique for embedded system implementation

“…According to Figure A the statistics of estimation error of fundamental amplitude have little change after 1000 runs. This results in sufficient information for evaluating final values of uncertain parameters . Figure B shows the results of MC runs for fundamental amplitude estimation.…”

“…A total of 8 uncertain parameters should be set for H∞-CKF. Selected parameters for CKF include 6 values, 5 diagonal elements of model noise matrix Q, and 1 element of covariance matrix of measurement noise m. As well, covariance matrix of measurement noise R and tuning factor α 26 Figure 2B shows the results of MC runs for fundamental amplitude estimation. As well, the convergence characteristics of the proposed method are depicted in Figure 2.…”

A Monte Carlo-Based Iterative Extended Kalman Filter for Bearings-Only Tracking of Sea Targets

“…The term "Monte Carlo method" is used to embrace a wide range of problem-solving techniques which use random numbers in input to produce the output statistics [32][33][34]. Examples of MC methods in science are classical MC (drawing samples from distributions to determine the desired output), quantum MC (used in physics), and simulation MC (algorithms that evolve configurations depending on predefined models).…”

Incorporating irregular nonlinear waves in simulation of dropped cylindrical objects

“…Development of equations of motion of a dropped object into calm water summarized in the following is mainly based on the previous works. 1,2,[10][11][12][13] To consider the effects of currents and/or waves, the modified equations of motion for a dropped object are expressed as follows The above equations are solved in time domain. For each time step, translational and rotational motions in local coordinate system (oxyz) can be obtained.…”

“…Forces and moments tangential to cylinder axis. The drag force tangential to the cylinder axis (x-axis) is represented by equation (13). The first term in equation 13is the frictional drag term obtained from boundary layer theory 17 and the second term denotes a form drag component 18…”