An Open-Source Test Environment for Effective Development of MARG-Based Algorithms

Research on Gradient-Descent Extended Kalman Attitude Estimation Method for Low-Cost MARG

Aiming at the problem of the weak dynamic performance of the gradient descent method in the attitude and heading reference system, the susceptibility to the interference of accelerometers and magnetometers, and the complex calculation of the nonlinear Kalman Filter method, an extended Kalman filter suitable for a low-cost magnetic, angular rate, and gravity (MARG) sensor system is proposed. The method proposed in this paper is a combination of a two-stage gradient descent algorithm and the extended Kalman filter (GDEKF). First, the accelerometer and magnetometer are used to correct the attitude angle according to the two-stage gradient descent algorithm. The obtained attitude quaternion is combined with the gyroscope measurement value as the observation vector of EKF and the calculated attitude of the gyroscope and the bias of the gyroscope are corrected. The elimination of the bias of the gyroscope can further improve the stability of the attitude observation results. Finally, the MARG sensor system was designed for mathematical model simulation and hardware-in-the-loop simulation to verify the performance of the filter. The results show that compared with the gradient descent method, it has better anti-interference performance and dynamic performance, and better measurement accuracy than the extended Kalman filter.

Autonomous Driving Control Based on the Perception of a Lidar Sensor and Odometer

Recently, many artificial intelligence applications in smart cars have been utilized in real life. Making an unmanned ground vehicle (UGV) capable of moving autonomously has become a critical topic. Hence, in this work, a novel method for a UGV to realize path planning and obstacle avoidance is presented using a deep deterministic policy gradient approach (DDPG). More specifically, the lidar sensor mounted on the vehicle is utilized to measure the distance between the vehicle and the surrounding obstacles, and the odometer measures the mileage of the vehicle for the purpose of estimating the current location. Then, the above sensed data are treated as the training data for the DDPG training procedure, and several experiments are performed in different settings utilizing the robot operating system (ROS) and the Gazebo simulator with a real robot module, TurtleBot3, to present a comprehensive discussion. The simulation results show that using the presented design and reward architecture, the DDPG method is better than the classic deep Q-network (DQN) method, e.g., taking fewer steps to reach the goal, less training time to find the smallest number of steps for reaching the goal, and so on.

Spatial Models and Neural Network for Identifying Sustainable Transportation Projects with Study Case in Querétaro, an Intermediate Mexican City

The construction of urban and transport indicators aims for a better diagnosis that enables technical and precise decision-making for the public administration or private investment. Therefore, it is common to make comparisons and observe which has better diagnosis results in a diversity of indexes and models. The present study made a comparative analysis of spatial models using artificial intelligence to estimate transport demand. To achieve this goal, the audit field was recollected in specific urban corridors to measure the indicators. A study case in Querétaro, an emergent city in the Mexican region known as El Bajío, is conducted. Two similar urban avenues in width and length and close to each other were selected to apply a group of spatial models, evaluating the avenues by segments and predicting the public transport demand. The resulting database was analyzed using Artificial Neural Networks. It displays specific indicators that have around 80% of correlations. The results facilitate the localization of the avenue segments with the most volume of activity, supporting interventions in urban renewal and sustainable transportation projects.

Neural Network and Spatial Model to Estimate Sustainable Transport Demand in an Extensive Metropolitan Area

Urban renewal projects worldwide focus mainly on resolving motorized, personal, and low occupancy problems instead of sustainable mobility. As part of the process, traditional field audits have a high cost in time and resources. This paper reviews a spatial model of accessibility and habitability of the streets, oriented to the location of the volume of people moving sustainably out of an extensive street network. The exercise site is in the Monterrey Metropolitan Area, the second largest in Mexico. Here, the population that moves sustainably as the collective (public and enterprise transportation) and the active (cycling, walking, and others) represents a considerable portion (49%) of travelers, thus, confirming the need for intervention. The spatial model is elaborated in a Geographical Information System (GIS), and the main results are compared with the actual public transport demand using a neural networks process. The results of the tool as a predictor have a 91% efficiency, making it possible to determine the location of urban renewal projects related to the volume of people moving sustainably.

The Impacts of COVID-19 on Technological and Polytechnic University Teachers

University teachers have adapted to different situations during the development of distance learning due to the pandemic caused by the COVID-19 virus. This study was conducted by assigning a data collection instrument to 993 teachers who are part of 15 technological universities (TUs) and 7 polytechnic universities (PUs) to determine how they were affected by COVID-19. The questions asked were related to the social, economic, academic, emotional, and health effects experienced. The results show that 63% of the teachers working online complained that online teaching invaded their family privacy; 56% pointed out that working from home and the virtual classes affected their performance as teachers; 90% of the teachers thought that they dedicated too much extra to preparing for their classes; 15% were stressed; 4% felt negative under the new teaching scheme of virtual classes; finally, 38% of the teachers stated that repeated interaction with electronic devices had a lot of negative impacts on their emotional wellbeing. By means of a G-test, it was determined that gender was independent from the studied effects. Through a multiple correspondence analysis (MCA), it was determined that, of the total number of teachers who responded to the questionnaire, half were comfortable with the online teaching model and the other half were not. The most impacted effects were the economic, training and connectivity independently from the gender.

Sensorless Estimation Based on Neural Networks Trained with the Dynamic Response Points

In the present work, a neuronal dynamic response prediction system is shown to estimate the response of multiple systems remotely without sensors. For this, a set of Neural Networks and the response to the step of a stable system is used. Six basic characteristics of the dynamic response were extracted and used to calculate a Transfer Function equivalent to the dynamic model. A database with 1,500,000 data points was created to train the network system with the basic characteristics of the dynamic response and the Transfer Function that causes it. The contribution of this work lies in the use of Neural Network systems to estimate the behavior of any stable system, which has multiple advantages compared to typical linear regression techniques since, although the training process is offline, the estimation can perform in real time. The results show an average 2% MSE error for the set of networks. In addition, the system was tested with physical systems to observe the performance with practical examples, achieving a precise estimation of the output with an error of less than 1% for simulated systems and high performance in real signals with the typical noise associated due to the acquisition system.

Estimation of Transfer Function Coefficients for Second-Order Systems via Metaheuristic Algorithms

The present research develops the parametric estimation of a second-order transfer function in its standard form, employing metaheuristic algorithms. For the estimation, the step response with a known amplitude is used. The main contribution of this research is a general method for obtaining a second-order transfer function for any order stable systems via metaheuristic algorithms. Additionally, the Final Value Theorem is used as a restriction to improve the velocity search. The tests show three advantages in using the method proposed in this work concerning similar research and the exact estimation method. The first advantage is that using the Final Value Theorem accelerates the convergence of the metaheuristic algorithms, reducing the error by up to 10 times in the first iterations. The second advantage is that, unlike the analytical method, it is unnecessary to estimate the type of damping that the system has. Finally, the proposed method is adapted to systems of different orders, managing to calculate second-order transfer functions equivalent to higher and lower orders. Response signals to the step of systems of an electrical, mechanical and electromechanical nature were used. In addition, tests were carried out with simulated signals and real signals to observe the behavior of the proposed method. In all cases, transfer functions were obtained to estimate the behavior of the system in a precise way before changes in the input. In all tests, it was shown that the use of the Final Value Theorem presents advantages compared to the use of algorithms without restrictions. Finally, it was revealed that the Gray Wolf Algorithm has a better performance for parametric estimation compared to the Jaya algorithm with an error up to 50% lower.

An Open-Source Test Environment for Effective Development of MARG-Based Algorithms

This paper presents an open-source environment for development, tuning, and performance evaluation of magnetic, angular rate, and gravity-based (MARG-based) filters, such as pose estimators and classification algorithms. The environment is available in both ROS/Gazebo and MATLAB/Simulink, and it contains a six-degrees of freedom (6 DOF) test bench, which simultaneously moves and rotates an MARG unit in the three-dimensional (3D) space. As the quality of MARG-based estimation becomes crucial in dynamic situations, the proposed test platform intends to simulate different accelerating and vibrating circumstances, along with realistic magnetic perturbation events. Moreover, the simultaneous acquisition of both the real pose states (ground truth) and raw sensor data is supported during these simulated system behaviors. As a result, the test environment executes the desired mixture of static and dynamic system conditions, and the provided database fosters the effective analysis of sensor fusion algorithms. The paper systematically describes the structure of the proposed test platform, from mechanical properties, over mathematical modeling and joint controller synthesis, to implementation results. Additionally, a case study is presented of the tuning of popular attitude estimation algorithms to highlight the advantages of the developed open-source environment.