The most common methods to improve the quality of images with insufficient visibility are retinex-based and gamma correction methods. The fundamental assumption of retinex theory is that the color of an object can be represented as the multiplication of its illumination and reflectance. The retinex-based method improves the quality of the insufficiently visible image by repairing its illumination. The multi-scale retinex (MSR) is a classic retinex-based method. Though MSR better enhances the details of the image, it sometimes reverses its lightness value. The method based on adaptive gamma correction with weighting distribution (AGCWD) is to modify the visibility of images by gamma function. However, AGCWD provides a good enhancement effect on low-contrast areas, it also enhances the high-light region making it too bright. In this paper, a method that combines the advantages of MSR and AGCWD is proposed. Firstly, the advatages of MSR and AGCWD are preserved into detailed image through the weight that considers illumination. Then, the image constructed by combining the detailed and original images could maintain the contrast of the high-light region and enhance details of the low-light region. The validity of the proposed method is shown by experiments using several images.

This study presents a generalized Tsallis entropy-based fuzzy c-means (GTFCM) clustering algorithm. Furthermore, the results of this study show that the behavior of GTFCM, at an infinity point of the fuzzy classification function, is similar to that of some conventional clustering algorithms. This result implies that such behavior is determined by a certain part of the GTFCM objective function.

The external reward plays an important role in the reinforcement learning process, and the quality of its design determines the final effect of the algorithm. However, in several real-world scenarios, rewards extrinsic to the agent are extremely sparse. This is particularly evident in mobile robot navigation. To solve this problem, this paper proposes a curiosity-based autonomous navigation algorithm that consists of a reinforcement learning framework and curiosity system. The curiosity system consists of three parts: prediction network, associative memory network, and curiosity rewards. The prediction network predicts the next state. An associative memory network was used to represent the world. Based on the associative memory network, an inference algorithm and distance calibration algorithm were designed. Curiosity rewards were combined with extrinsic rewards as complementary inputs to the Q-learning algorithm. The simulation results show that the algorithm helps the agent reduce repeated exploration of the environment during autonomous navigation. The algorithm also exhibits a better convergence effect.

Due to the advent of the COVID-19 pandemic, the Philippine government encouraged enterprises and businesses to utilize flexible work arrangements such as work-from-home (WFH) or telecommuting setup. Nowadays, the key components necessary for a telecommuting include a WiFi-enabled IT equipment, secured work environment, and reliable internet connection, while research shows that type of work and computer literacy are also key factors for telework implementation. Multiple studies in relation to telework have already been conducted but some studies were deemed inconclusive and need further analysis. Therefore, in this study, a Mamdani fuzzy-based model was developed for telework capability assessment for Philippine government employees based on four significant factors namely: internet speed, IT equipment availability, computer literacy, and type of work, which are expressed in linguistic representations. The proposed fuzzy system can provide a feedback telework capability score based on the four input parameters which may also be characterized with the potential telecommuting cost requirement.

Identification of fungi infecting Zea mays leaves and sub-classifying them to have correct course management in the earlier stages is lucrative. To develop a nondestructive and low-cost classification model of corn leaves infected by Setosphaeria turcica (ST), Cercospora zeae-maydis (CZM), and Puccinia sorghi (PS) fungi using image filtering and transfer learning model. Corn leaf images were categorized based on fungal-infection and stored in an image library. All images were then processed to show different intensities and then utilized to filter the images. An original RGB-based CNN model has been compared with selected pre-trained models of VGG16 and EfficientNet-b0 with inputs of both unfiltered and filtered RGB images. Results showed that the EfficientNet-b0 with filtered images model (fMaize) exhibited the highest accuracy of 97.63%, sensitivity of 97.99%, specificity of 97.38, quality index of 97.68%, and F-score of 96.48%. Consequently, the experimental results revealed that deep transfer learning models fed with filtered images produced higher accuracy than models that simply employed RGB images. Thus, transfer learning was proven to be a valuable tool in enhancing CNN image classification accuracy.

Deep learning has attracted attention widely as the successful application of deep learning for vision tasks, such as image classification, object detection and so on. Due to the robustness and universality of deep learning, automotive manufacturing, a crucial part of national economy, needs deep learning to make production lines more intelligent and improve efficiency. However, some superior generally deep learning models, such as ViT, TNT, and Swin transformer, cannot meet automotive manufacturing requirements with high accuracy on a specific scene. As for automotive production lines, engineers usually adopt some smart designs, which can provide prior knowledge for designing deep learning models. Specifically, in an image, the position of target is usually fixed. Therefore, in order to take advantage of prior position, this paper designs a local mixer with prior position to capture local feature. Its main idea is that dividing the whole feature map into window feature maps and connecting window feature maps along channel dimension in order to make convolution kernel parameters for each window feature map are independent from others. Besides, MLP is adopted as global mixer to capture global feature and the pyramidal architecture with CNN is adopted. Comprehensive results demonstrate the effectiveness of proposed model on cars’ type recognition. In particular, the proposed model achieves 97.938% accuracy on our data set, surpassing some transformer-like models.

Poultry, like quails, is sensitive to stressful environments. Too much stress can adversely affect birds’ health, causing meat quality, egg production, and reproduction to degrade. Posture and behavioral activities can be indicators of poultry wellness and health condition. Animal welfare is one of the aims of precision livestock farming. Computer vision, with its real-time, non-invasive, and accurate monitoring capability, and its ability to obtain a myriad of information, is best for livestock monitoring. This paper introduces a quail detection mechanism based on computer vision and deep learning using YOLOv5 and Detectron2 (Faster R-CNN) models. An RGB camera installed 3 ft above the quail cages was used for video recording. The annotation was done in MATLAB video labeler using the temporal interpolator algorithm. 898 ground truth images were extracted from the annotated videos. Augmentation of images by change of orientation, noise addition, manipulating hue, saturation, and brightness was performed in Roboflow. Training, validation, and testing of the models were done in Google Colab. The YOLOv5 and Detectron2 reached average precision (AP) of 85.07 and 67.15, respectively. Both models performed satisfactorily in detecting quails in different backgrounds and lighting conditions.

Water quality is crucial for maintaining a sustainable living environment in aquaculture. Limnological parameters affects the fish physiology, growth rate, and feed efficiency and may lead to high mortality rate under extreme conditions. The development of an adaptive aquaculture monitoring system for water quality using fuzzy logic will address this problem. Using Mamdani-type fuzzy inferences system (FIS) model, the input limnological parameters such as pH, temperature, total dissolved solids, and dissolved oxygen levels were transformed to four output states: excellent, good, poor, and toxic, for the prediction of water quality. For the simulation and evaluation of the developed FIS, MATLAB Simulink was used. Results of this study can be integrated with a feedback system for appropriate treatments including filtering, aeration, and water flushing to maintain safe environment for Nile tilapia.

To enhance trajectory tracking performance of atomic force microscope system, a two-degree of freedom fractional order PID (2-DOF FOPID) control approach based on back propagation (BP) neural network is proposed in this paper. At first, principle and structure of the proposed control approach is presented. Then, 2-DOF FOPID controller is designed, including in feedforward and feedback controller, fractional calculus and approximation of fractional operator. Meanwhile, the parameters of controller are analyzed. Based on them, a BP neural network is built to adjust the parameters in this control structure according to the error between the reference trajectory and the actual output. Finally, the proposed control approach is conducted in atomic force microscope tracking control experiment, experimental results verify the effectiveness and improvement of the proposed control approach.

The experimental studies on prototypes printed in 3D with polylactic acid (PLA) material still seek to characterize the mechanical behavior and the deformations of these printed samples according to the various solicitations. The huge number of parameters intervening in these properties makes the control of process difficult and expensive. Previous studies on the impact of these parameters on the mechanical properties are limited to the investigation of a very less number of parameters. The objective of the present study is to take advantage of artificial intelligence tools, and to exploit the experimental results, in order to present artificial models that are able to optimize the choice of parameters intervening in the properties (tensile strength) of printed parts.

The vehicle routing problem (VRP) has many applications in goods distribution and goods transportation. Today, many companies have requirements for VRP with multiple pickup and multiple delivery within due time. This problem is called multiple pickup and multiple delivery vehicle routing problem with time window (PDPTW). PDPTW has many constraints and ant colony optimization (ACO) has been used to solve it although ACO creates too many infeasible routes. Moreover, it often gets trapped in local optimum. To solve these problems, this paper proposed an improved ACO by using the route elimination technique and the pheromone reset technique. The ACO with route elimination technique, it has proven to solve the PDPTW problem with increased performance. The proposed technique was tested on datasets from the Li & Lim’s PDPTW benchmark problems and provided more satisfactory results compared to other ACO techniques.

The accurate identification and characterization of the accelerometer dynamic model parameters play an important role in improving the dynamic performance of the device or system with an accelerometer. To overcome the problem that the traditional single degree of freedom (SDF) dynamic model of the accelerometer cannot describe the dynamic characteristics beyond the first resonant frequency of the accelerometer, a two degree of freedom (TDF) dynamic model of the accelerometer was constructed. On this basis, a parameter identification method for the TDF dynamic model of the accelerometer based on the feature points coordinate estimation and amplitude correction was proposed. First, the zero frequency point coordinates of the accelerometer frequency response were obtained by the H_v method. The first and second resonance point coordinates were estimated by discrete spectrum correction and the least square (DSC-LS) method. Then, the amplitude correction coefficient was applied to eliminate the influence of series coupling on the amplitude. Finally, the TDF dynamic model parameters of the accelerometer were calculated through the feature point coordinates. The experimental results show that the method has high accuracy and can avoid the influence of series coupling on the parameter identification accuracy of the accelerometer’s TDF dynamic model without complex derivation and decoupling operations. The identified TDF dynamic model of the accelerometer can represent the dynamic characteristics with a higher frequency range.

In this study, the dynamic joint scheduling problem for processing machines and transportation robots in a flexible job shop is investigated. The study aims to minimize the order completion time (makespan) of a job shop manufacturing system. Considering breakdowns, order insertion and battery charging maintenance of robots, an event-driven global rescheduling strategy is adopted. A novel memetic algorithm combining genetic algorithm and variable neighborhood search is designed to handle dynamic events and obtain a new scheduling plan. Finally, numerical experiments are conducted to test the effect of the improved operators. For successive multiple rescheduling, the effectiveness of the proposed algorithm is verified by comparing it with three other algorithms under dynamic events, and through statistical analysis, the results verify the effectiveness of the proposed algorithm.

Route recommendation continues to manifest noteworthy contributions to the intelligent transportation system field of research as it evolves through time. Early related studies helped passengers and tourists experience a more convenient travel. At the same time, these helped transport planners analyze people’s trip preferences and its correlation with the region-specific economic status in a more time-relevant data. Majority, however, require historical data and heavy data collection methods. For user quantified metrics such as route cost in terms of travel time and distance, the complexity and sparsity of preferences between travelers are persistent challenges. The strategic transit route recommendation proposed in this study takes into account multiple trip features (both quantitative and qualitative) desirability using logit model and the optimal travel time with respect to a given road traffic condition, headway, and passenger demand. The chosen area of study is the Western Visayas region of the Philippines specific to the public utility bus (PUB) and jeepney (PUJ) transit routes. The results of the research exhibited the feasibility of an optimal and strategic recommendation of public transportation route for passengers considering present time relevant trip conditions rather than relying on the historical data which are difficult to obtain, or worse, non-existent.

In this work, we study the application of multimodal analogical reasoning to image retrieval. Multimodal analogy questions are given in a form of tuples of words and images, e.g., “cat”:“dog”::[an image of a cat sitting on a bench]:?, to search for an image of a dog sitting on a bench. Retrieving desired images given these tuples can be seen as a task of finding images whose relation between the query image is close to that of query words. One way to achieve the task is building a common vector space that exhibits analogical regularities. To learn such an embedding, we propose a quadruple neural network called multimodal siamese network. The network consists of recurrent neural networks and convolutional neural networks based on the siamese architecture. We also introduce an effective procedure to generate analogy examples from an image-caption dataset for training of our network. In our experiments, we test our model on analogy-based image retrieval tasks. The results show that our method outperforms the previous work in qualitative evaluation.

3D point cloud semantic segmentation has been widely used in industrial scenes and has attracted continuous attention as a critical technology for understanding the intelligent robot scene. However, extracting visual semantics in complex environments remains a challenge. We propose the Seg-PointNet model based on multi-layer residual structure and feature pyramid for the LiDAR point cloud data semantic segmentation task in the complex substations scene. The model is based on the PointNet network and introduces a multi-scale residual structure. The residual structure multilayer perception (RES-MLP) model is proposed to fully excavate features at different levels and improve the characterization capabilities of complex features. Moreover, the 3D point cloud feature pyramid module is proposed to characterize the substation scene’s semantic features. We tested and verified the Seg-PointNet model on a self-built substation cloud point (SCP) dataset. The results show that the proposed Seg-PointNet model effectively improves the point cloud data segmentation accuracy, with an accuracy of 89.23% and mean intersection over union (mIoU) of 63.57%. This shows that the model can be applied to substation scenarios and provide technical support to intelligent robots in complex substation environments.

At present, the world faces the challenge of making education available to all the people of the world, especially in under-developed areas, where there is a lack of human resources. To improve educational standards and address this lack of human resources, we believe that robots can be a good teaching aid, to assist both the teacher and the student. This study investigated the entrainment phenomenon between human interactions with a robot as a third-party influence to provide learning packages that will be suited for students and will reduce the strain on teachers. We used dynamic time warping (DTW) algorithm in our time series analysis to effectively calculate the similarity of the brainwaves of the subjects. Then, we defined the entrainment phenomenon between the subjects. To ensure the complexity of this experiment, crosswords were used for this task, which subjects were required to complete in cooperation with each another. At different points of time, the robot would use different expressions, actions, and words as reminders or hints, aiding communication and better cooperation between subjects. The timing of the robotic reminders, synchronization, conversations between the subjects in the crossword tasks, and the completion rate of the crossword puzzles were recorded. The average completion rate improved by 10% with the aid of the robot, as opposed to not using the robot. The results of this study prove that robotic participation in human interaction is beneficial, and that implementing robotic assistance will improve education.

In recent years, the topic of reducing fuel consumption and greenhouse gas emission has become one of the major focuses on the automotive industry leading toward the development of electric vehicles to create awareness of environmental protection. Thus, the development of hybrid electric vehicle (HEV), plug-in hybrid electric vehicle (PHEV), and fully electrical vehicle (EV) has started growing up to replace the gasoline car, which is fully depends on fuel to operate, to help fight against the world climate change issues. This research is mainly focused on solving the problem of charging period of traditional used batteries pack, energy storage system of EV, and the limitation on travel distance for EV with the use of batteries pack as an energy source. The proportional-integral (PI) controller based on particle swarm optimization (PSO) algorithm is implemented in this simulation to optimize the speed of BLDC motor by obtaining an optimized parameter of K_p and K_i. The MATLAB/Simulink software is used for graphical modelling, simulating, and analyzing the behavior of supercapacitor in various condition. The simulation results represent the proposed PSO-based energy management method can achieve greater energy efficiency as compared to the traditional method. All in all, moving forward in developing a fully electric buses or vehicles can bring society into a new generation of zero greenhouse gas emissions. In this paper, the optimization of the PI controller based on PSO algorithm is applied and the results show that there is an increment of 6% in total distance traveled by the EV. Besides, there is the 3.69% of improvement for maximum speed and peak to peak speed of the EV and 14.57% of improvement in terms of average speed of EV within the total travel duration of 1300 s.

Option pricing plays an important role in modern finance. This paper investigates the uncertain option pricing problems based on uncertainty theory by using the method to calculate the optimistic value of uncertain returns of options instead of the method of traditional expected value in the sense of the weighted average. The pricing formulas of the European and American options are derived for Liu’s uncertain stock model and Peng’s mean-reverting stock model which are two basic and representative uncertain stock models in uncertain finance. In the end, some numerical experiments are given to illustrate the effectiveness of the obtained results.

The dominant position in crude oil pricing affects a country’s energy and economic security. We construct different time-series forecasting models to analyze the long-term trends of Shanghai crude oil futures and guide current price trading. First, we built a classic ARIMA model as the benchmark. It is a good choice for short-term investors but does not apply to long-term trend prediction. We propose an improved HW-Prophet model that can maximize its advantages in selecting and reconstructing mutation points and reduce the forecast error by approximately 22%. The Prophet model is then introduced into the prediction of futures price series, and the Haar wavelet and KS (HWKS) algorithm and sliding window are used to optimize the change-point selection of the Prophet model. Finally, we created different combinations of long short-term memory (LSTM) and HW-Prophet models based on residual correction and optimal weighting. The results showed that the combined residual correction method was unsuitable for the two single models in this study. Because the residual sequence predicted by the HW-Prophet model is similar to white noise, it is difficult for LSTM to learn the effective dynamic laws from it. By minimizing the error sum of squares to solve the optimal weight coefficients and combining the two models linearly, the prediction advantages of a single model in different intervals can be effectively transmitted, and the prediction accuracy is greatly improved. The HW-Prophet-LSTM combination model constructed in this study based on the optimal weight is currently one of the best models for predicting the long-term trends of Shanghai crude oil futures.

In this paper, we propose an empathy method for communicating robots, aiming to realize robots with human-like empathy. The proposed method determines whether the robot empathizes with humans by obtaining an empathy coefficient from the robot’s own emotions and the estimated human emotions. Weiner’s empathy experiment with a sick person showed that the robot exhibited an internal state similar to that of characters inferred from the scenario. In addition, we conducted an impression evaluation experiment on the robot’s response with and without empathy and found a significant difference at the 5% level of significance in the Mann–Whitney U test. Therefore, the effectiveness of the proposed method is suggested.

Processing ultrasonic echo signals to obtain high-precision residual thickness information of the pipeline wall is the key to nondestructive testing of corrosion of a long-distance pipeline. The traditional power spectrum estimation method assumes that an analyzed echo signal is Gaussian, and the useful information is insufficiently extracted, which leads to errors in the processing results. In this paper, to solve this problem, the bispectrum, which requires the least amount of computation in higher-order spectral estimation, is proposed to process an echo signal with a non-minimum phase and non-Gaussian characteristics. The bispectrum is projected onto a one-dimensional frequency space using the dimensionality reduction method, and one-dimensional diagonal slices of the bispectrum are extracted to analyze the characteristics of the echo signal, which significantly improves the intuitiveness of data processing. The experimental results show that the bispectrum dimensionality reduction method has high accuracy in processing ultrasonic echo signals, and the relative error of the residua wall thickness is below 2%. A C-scan image displaying the shape, size, depth, and other characteristics of pipeline corrosion obtained by the proposed method is much better than that using the traditional power spectrum estimation method. Therefore, the proposed method is suitable for nondestructive testing of corrosion of long-distance pipelines.