Real-time route identification, mass estimation and velocity profile optimisation

This dataset contains the recorded test data and executable programme files to perform route identification, mass estimation and velocity profile optimisation in real-time. The software used includes Excel, Octave 4.1, Matlab 2021b and Simulink. All of these form part of a PhD thesis, each set contributing to a section of the work in isolation. In the folder named, Phase 1_underground navigation, there is all the test data for road tests, where magnetic heading, barometric altitude, driving torque etc. is stored. These are used to identify routes that were travelled based on patterns in heading and change in altitude. It was found that patterns in heading and barometric altitude can be effectively used to identify a known route by.
In the folder named, Phase 2 Mass estimation, the data available for driving torque, speed, incline etc., which is used to obtain an estimate for a vehicle's mass in real-time. By applying newton's second law for a vehicle on an incline, and rearranging the equations to solve for the mass, M, it is possible to obtain a real-time estimate for the vehicle's mass. In the folder named, Phase 3 optimisation, the codes required to run an optimisation of a velocity profile for a vehicle on a known route is presented. The initial guess plays a significant role in any optimisation problem, and in this study a method was developed that makes use of an inverted version of the route topographic profile as an initial guess to the optimisation problem. This was proven to be very effective, and also proved to be capable of being used as an estimate to the optimal problem, at a significantly reduced simulation time (20 seconds, vs 3 hours). This method was converted to a lookup table which is able to solve in milliseconds, allowing real-time application.

本数据集收录了实时进行路线识别、质量估计和速度曲线优化所需的测试记录数据和可执行程序文件。所使用的软件包括 Excel、Octave 4.1、Matlab 2021b 和 Simulink。所有这些软件均构成博士学位论文的一部分，各自独立贡献于论文的特定章节。在名为 'Phase 1_underground navigation' 的文件夹中，存储了道路测试的全部测试数据，其中包含磁方位、气压高度、驱动力矩等信息。这些数据用于根据航向和高度变化模式识别已行驶路线。研究发现，航向和气压高度的模式可以有效用于通过已知路线识别。在名为 'Phase 2 Mass estimation' 的文件夹中，提供了用于获取车辆质量实时估计的驱动力矩、速度、坡度等数据。通过应用牛顿第二定律于坡道上的车辆，并重新排列方程以求解质量 M，可以实现对车辆质量的实时估计。在名为 'Phase 3 optimisation' 的文件夹中，展示了用于对已知路线上的车辆速度曲线进行优化的代码。在优化问题中，初始猜测扮演着至关重要的角色，本研究开发了一种方法，该方法利用路线地形剖面图的倒置版本作为优化问题的初始猜测。这一方法已被证明非常有效，并且能够显著减少仿真时间（20秒，相较于3小时），同时证明了其作为最优问题估计的能力。该方法已被转换为一个查找表，能够在毫秒级别内解决问题，从而实现实时应用。