*A theoretical demonstration for reinforcement learning of PI control dynamics for optimal speed control of DC motors by using Twin Delay Deep Deterministic Policy Gradient Algorithm

Abstract

Abstract:To benefit from the advantages of Reinforcement Learning (RL) in industrial control applications, RL methods can be used for optimal tuning of the classical controllers based on the simulation scenarios of operating conditions. In this study, the Twin Delay Deep Deterministic (TD3) policy gradient method, which is an effective actor-critic RL strategy, is implemented to learn optimal Proportional Integral (PI) controller dynamics from a Direct Current (DC) motor speed control simulation environment. For this purpose, the PI controller dynamics are introduced to the actor-network by using the PI-based observer states from the control simulation environment. A suitable Simulink simulation environment is adapted to perform the training process of the TD3 algorithm. The actor-network learns the optimal PI controller dynamics by using the reward mechanism that implements the minimization of the optimal control objective function. A setpoint filter is used to describe the desired setpoint response, and step disturbance signals with random amplitude are incorporated in the simulation environment to improve disturbance rejection control skills with the help of experience based learning in the designed control simulation environment. When the training task is completed, the optimal PI controller coefficients are obtained from the weight coefficients of the actor-network. The performance of the optimal PI dynamics, which were learned by using the TD3 algorithm and Deep Deterministic Policy Gradient algorithm, are compared. Moreover, control performance improvement of this RL based PI controller tuning method (RL-PI) is demonstrated relative to performances of both integer and fractional order PI controllers that were tuned by using several popular metaheuristic optimization algorithms such as Genetic Algorithm, Particle Swarm Optimization, Grey Wolf Optimization and Differential Evolution.

Duyuru kategorisine gönderildi

*COVID-19 and human development: An approach for classification of HDI with deep CNN

Abstract

The measures taken during the pandemic have had lasting effects on people’s lives and perceptions of the ability of national and multilateral institutions to drive human development. Policies that changed people’s behavior were at the heart of containing the spread of the virus. As a result, it has become a systemic human development crisis affecting health, the economy, education, social life, and accumulated gains. This study shows how the relationship of the Human Development Index (HDI), which has combined effects on health, education, and the economy, should be considered in the context of pandemic factors. First, COVID-19 data of the countries received from a public and credible source were extracted and organized into an acceptable structure. Then, we applied statistical feature selection to determine which variables are closely related to HDI and enabled the Deep Convolutional Neural Network (DCNN) model to give more accurate results. The Continuous Wavelet Transform (CWT) and scalogram methods were used for the time-series data visualization. Three different images of each country are combined into a single image to penetrate each other for ease of processing. These images were made suitable for the input of the ResNet-50 network, which is a pre-trained DCNN model, by going through various preprocessing processes. After the training and validation processes, the feature vectors in the fc1000 layer of the network were drawn and given to the Support Vector Machine Classifier (SVMC) input. We achieved total performance metrics of specificity (88.2%), sensitivity (96.5%), precision (99%), F1 Score (94.9%) and MCC (85.9%).

Duyuru kategorisine gönderildi