The application of deep learning technology has increased rapidly in recent years. Technologies in deep learning increasingly emulate natural human abilities, such as knowledge learning, problem-solving, and decision-making. In general, deep learning can carry out self-training without repetitive programming by humans. Convolutional neural networks (CNNs) are deep learning algorithms commonly used in wide applications. CNN is often used for image classification, segmentation, object detection, video processing, natural language processing, and speech recognition. CNN has four layers: convolution layer, pooling layer, fully connected layer, and non-linear layer. The convolutional layer uses kernel filters to calculate the convolution of the input image by extracting the fundamental features. The pooling layer combines two successive convolutional layers. The third layer is the fully connected layer, commonly called the convolutional output layer. The activation function defines the output of a neural network, such as 'yes' or 'no'. The most common and popular CNN activation functions are Sigmoid, Tanh, ReLU, Leaky ReLU, Noisy ReLU, and Parametric Linear Units. The organization and function of the visual cortex greatly influence CNN architecture because it is designed to resemble the neuronal connections in the human brain. Some of the popular CNN architectures are LeNet, AlexNet and VGGNet.

The use of DC motors is now common because of its advantages and has become an important necessity in helping human activities. Generally, motor control is designed with PID control. The main problem that is often discussed in PID is parameter tuning, namely determining the value of the Kp, Ki, and Kd parameters in order to obtain optimal system performance. In this study, one method for tuning PID parameters on a DC motor will be used, namely the Particle Swarm Optimization (PSO) method. Parameter optimization using the PSO method has stable results compared to other methods. The results of tuning the PID controller parameters using the PSO method on the MATLAB Simulink obtained optimal results where the value of Kp = 8.9099, K = 2.1469, and Kd = 0.31952 with the value of rise time of 0.0740, settling time of 0.1361 and overshoot of 0. Then the results of hardware testing by entering the PID value in the Arduino IDE software produce a stable motor speed response where Kp = 1.4551, Ki= 1.3079, and Kd = 0.80271 with a rise time value of 4.3296, settling time of 7.3333 and overshoot of 1.

Flight control design of unmanned aerial vehicles UAVs is becoming increasingly important due to advances in computational power of computers with lower cost. The control algorithms are mainly employed for the attitude and position control of the UAVs. In the past decades, quadrotors have become the most popular UAVs, their adaptability and small size. They are employed to carry out tasks such as delivery, exploration, fumigation, mapping, surveillance, rescue mission, traffic monitoring, and so on. While carrying out these tasks, quadrotor UAVs face various challenges, such as environmental disturbances, obstacles, and parametric and non-parametric perturbations. Therefore, they require robust and effective control to stabilize them and enhance their performance. This paper provides a survey of recent developments in control algorithms applied to attitude and position loops of quadrotor UAVs. In addition, the limitations of the previous control approaches are presented. In order to overcome the relative drawbacks of the previous control techniques and enhance the performance of the quadrotor, researchers are combining various control approaches to obtain the hybrid control architecture. In this study, a review of the recent hybrid control schemes is presented.

Noise Reduction in the Accelerometer and Gyroscope Sensor with the Kalman Filter Algorithm

Control of DC Motor Using Integral State Feedback and Comparison with PID: Simulation and Arduino Implementation

Electronic Nose (E-Nose) technology unlocks the fascinating world of electronic detection, identification, and analysis of scents and odors, paving the way for innovative research and promising applications. E-Nose mimics the human sense of smell and has gained significant attention and is applied in various fields, including the food, health and drug industries, safety and crime, and the environmental and agricultural sectors. This technology has the potential to improve quality control, medical diagnostics, and hazardous material detection processes. The E-Nose consists of a combination of gas sensors that mimic the olfactory receptors of the human nose. These sensors detect and respond to different scent molecules, resulting in unique response patterns that can be interpreted and analyzed. E-Nose has found application in the food industry to assess food quality, detect contamination, and monitor fermentation processes. In the health field, it has been used for disease diagnosis, monitoring patient health, and detecting cancerous tissue. In addition, E-Nose has been used for security purposes, such as detection of explosives and prohibited substances, as well as identification of counterfeit products. In addition, it has been used in environmental monitoring for air quality assessment and agriculture for disease detection in crops. Despite its promising potential, widespread adoption of E-Nose faces challenges related to sensor sensitivity, data analysis algorithms (complex data interpretation), response diversity, regulatory considerations, implementation complexity, and cost. This article reviews the latest developments in E-Nose technology, explores its applications and future potential, and highlights challenges that need to be addressed. This is considered important because E-Nose opens up a world of electronic scent identification, and analysis with the potential to improve quality control, diagnosis, and detection.