https://doi.org/10.5573/IEIESPC.2024.13.3.197

(Wenna Li);(Shuai Tang)

The assembly of various components in automobiles is an essential component of manufactured products, and currently, most of the assembly is done manually, which has high labor intensity and specific human errors. An unmanned workshop and intelligent processing were achieved by improving the vector control of assembly motors operated by a microcontroller with embedded wireless modules in the background. This study also performed two-bit spatial modeling of the workshop and then combined sensor technology to build the environment and transport this motor to the operating table. Therefore, the ant colony algorithm is used for path planning to achieve obstacle avoidance and smooth transportation of automatic transportation in the workshop. The torque and current pulse output using the embedded vector controller were relatively smooth, with small fluctuations up and down and current fluctuations within 1A. Hence, the controller can achieve vector control while maintaining a stable torque output. The studied vector controller can achieve a given torque output at 0.1s, highlighting the high efficiency of the method. The path planning results of the three types of obstacle transportation robots can achieve path smoothness, with an error of only about 0.12, demonstrating the excellent performance of this method. Several vector controllers satisfied the frame and seat assembly when used with the transport robot.

https://doi.org/10.5573/IEIESPC.2024.13.3.209

(Guiheng Zhi)

In recent years with continuous development of the computer vision field, there has been an increasing demand for fast and accurate recognition of human movement, especially in sports. This paper researches ballet movements, which are recognized and analyzed using a convolutional neural network (CNN) based on deep learning. Training of the CNN is improved by particle swarm optimization (PSO). Then, 1,000 ballet videos are used as a dataset to compare optimized CNN, traditional CNN, and support vector machine (SVM) methods. The results show that the improved CNN converged fastest, stabilizing after about five iterations, whereas the traditional CNN method took approximately 20 iterations to stabilize. Additionally, after convergence, error in the improved CNN was smaller than from the traditional CNN. The average recognition accuracy of the SVM method was 84.17%, with a recognition time of 3.32 seconds; for the traditional CNN method, it was 90.16% with a recognition time of 2.68 s; and for the improved CNN method, it was 95.66% with a recognition time of only 1.35 s.

https://doi.org/10.5573/IEIESPC.2024.13.3.215

(Vikram Gawali);(Milind Pande);(Munir Sayyad);(Raghunath Bhadade)

This paper proposes a comprehensive method for developing an automated drainage monitoring system using sensor networks, an autonomous underwater vehicle (AUV), and a network control center (NCC). The system collects sensor data, communicates with the NCC, and employs the AUV to identify and remove obstructions while monitoring its operation. The system performance evaluation, 3D drainage system modeling, and algorithm functionality revealed the effectiveness of the proposed methodology. These findings highlight the improved maintenance, sanitation, and reduced waterborne disease risks in urban areas. The Indian government's focus on developing "smart cities" and the challenges of poor subterranean drainage surveillance leading to water contamination and diseases has emphasized the importance of drainage systems. These findings highlight the efficacy and feasibility of the system for improving maintenance, sanitation, and reducing waterborne disease risks in urban and metropolitan areas.

https://doi.org/10.5573/IEIESPC.2024.13.3.225

(Zongkui Wang)

Community education is confronted with the challenge of managing information overload. The issue of prompt and precise course recommendations was addressed by proposing a community education course recommendation model based on a listwise learning algorithm for social network recommendation ranking and a dynamic community education course recommendation model based on trust relationships. The list-learning algorithm for social network recommendation ranking had a minimum and maximum cumulative gain of approximately 0.73 and 0.82, respectively. The dynamic recommendation model for community education courses based on trust relationships had a maximum coverage of approximately 0.75, with a root mean square error of 1.25?1.4. The number of dynamic recommendations varied between approximately 2400 and 4100. These methods outperformed other algorithms, suggesting that the model positively affects community education programs and promotes their growth. In summary, the constructed model showed favorable application effects in recommending community education curricula and promoting the development of community education.

https://doi.org/10.5573/IEIESPC.2024.13.3.236

(Vasily Sachnev);(Mahanand Belathur Suresh)

In this paper, we propose an automatic diagnostic tool for autism based on machine learning and structural MRI. A set of 989 relevant features extracted from structural magnetic resonance imaging (MRI) present in the ABIDE database was used. The tool has two steps: searching for the best set of features using a tailored binary coded genetic algorithm and then using the selected set of features to train a classifier using an extreme learning machine. The trained classifier can efficiently identify autism spectrum disorder (ASD) versus healthy controls. The proposed tailored binary coded genetic algorithm uses a statistical selective crossover designed specifically for the classification problem, which significantly speeds up the genetic algorithm and improves the classification of ASD versus healthy controls. Extensive experiments with the proposed automatic autism diagnostic tool clearly indicate the advantages of the proposed method compared to previous approaches.

https://doi.org/10.5573/IEIESPC.2024.13.3.243

(Hongtao Wang)

With the continuous development of internet information technology, cloud-computing task-scheduling platform technology is gradually maturing. Cloud computing is profoundly changing every aspect of people's lives and providing many conveniences. With the application of cloud computing in more fields, more extensive applications and efficient task scheduling algorithms have become increasingly important. This research focuses on the problem of task-scheduling methods for cloud computing platforms in customer-oriented online training systems. Based on the optimization of the ant colony algorithm, an ant colony optimization (ACO) cloud-computing task-scheduling algorithm is proposed. The research results indicate that when the number of tasks is 300, the makespan value of the optimized ant colony cloud scheduling algorithm (OACC) is 340, that of the discrete firefly algorithm (DFA) is 350, that of multi-objective differential evolution (MODE) is 380, and that of improved group search optimization (IGSO) is 409. The overall performance of OACC was 20.3% higher than that of IGSO. OACC maintained a low and stable degree of imbalance (DI) in different task count tests. At a task volume of 300, the overall utility evaluation of the ACO cloud-computing task-scheduling algorithm was 146, which is 31.5% higher than ACO, 18.7% higher than TACO, and 8.1% higher than LB-AACO. The experimental results meet expectations and indicate that the OACC cloud-computing task-scheduling algorithm proposed in the study has high task-processing ability and efficiency and is capable of scheduling tasks on cloud computing platforms for customer-oriented online training systems.

https://doi.org/10.5573/IEIESPC.2024.13.3.254

(Nandini Manickam);(Vijayakumar Ponnusamy)

In recent trends of growth in technologies, data management, maintenance of medical records, sharing of data, diagnosis of disease, and medication are the key areas where digital healthcare plays a vital role. Despite enormous improvement, handling huge amounts of data, privacy, secure sharing, accuracy, and computational speed remains challenging. Federated learning is a machine learning technology that allows distributed model training using users’ own data to train a model. The model update is done through a central server that aggregates individual users and sends a global model. This ensures privacy protection and is suitable for handling large data. Blockchain technology is a publicly distributed ledger that collects the information of nodes as blocks and sends a copy to all nodes in the network so that data transparency is maintained and secure. However, blockchain has a limitation in handling large volumes of data. In such cases, federated learning can be used with a blockchain for better performance. By integrating federated learning with blockchain, accurate prediction, computational speed, data security, privacy, and accuracy can be achieved. A comprehensive review of how various federated learning technologies can integrate with blockchain networks to achieve accuracy and efficiency is presented.

https://doi.org/10.5573/IEIESPC.2024.13.3.263

(Dengjin Li)

The optimal allocation of rural land use plays a unique role in developing rural industries. Therefore, realizing the effective use of land resources is the key to sustainable development. This research attempts to explore the modeling of rural land use in combination with an ant colony algorithm and multi-objective optimization problem under the background of the current digital economy and to explore land use and space allocation after optimal allocation. The experimental results showed that the multi-objective optimization model of land use proposed in this study could optimize the relevant objective functions so the entire optimization system can reach the optimal solution. The iterations of different objective functions under the three optimization models were compared. The four objective functions of carbon emission, minimum planning cost, adaptability value, and spatial agglomeration all iterated approximately 45 times. They began to converge under the premise of taking the land adaptability value and spatial agglomeration as optimization goals. The convergence rate was faster under this optimization model. In addition, the iteration and running time of the traditional ant colony algorithm, the genetic algorithm, and the improved ant colony algorithm under two different objective functions were compared.

https://doi.org/10.5573/IEIESPC.2024.13.3.273

(Wei Lou);(Rong Hu);(Gang Luo);(Rui Yang)

Power dispatch is an effective measure to achieve the rational utilization of power resources and reduce economic costs. On the other hand, existing power resource scheduling technologies have problems, such as low efficiency and inability to meet actual power demand. Therefore, search factors and selection strategies are introduced based on the preprocessing of power load data to improve the artificial bee colony (ABC) algorithm, solve this problem, and meet the power load demand. A power resource-scheduling model based on the ABC algorithm is constructed to achieve reasonable utilization of resources. According to the findings, the optimized method showed the best convergence performance among the six benchmark function tests. In the f2 function, it converged from 108. The minimum loss value was 104 when the function iteration value was 103. This suggests that the power resource scheduling technology based on improved methods has good application effects and can effectively achieve the reasonable allocation and utilization of power resources.

https://doi.org/10.5573/IEIESPC.2024.13.3.285

(Lianqin Zhu)

With the rapid development of society, the competition among various enterprises is also constantly strengthening. In view of this, research on applying various artificial intelligence methods to enterprise information management game systems came into being. In this study, principal component analysis reduces the dimensionality of each data method. Then, the fireworks algorithm improves and adjusts the numbers involved in back propagation. The Levenberg-Marquardt algorithm promotes training accuracy in the model, which jumps out of the local minimum and obtains the optimal solution. Research results show that when the number of iterations is set to 300, the model reaches the target error after only 110. The Fireworks Algorithm?Back Propagation (FWA-BP) test results indicate that while some samples had a prediction deviation of over 10, the variance between other models' predictions and the expected value was not significant, demonstrating a fitness degree of 0.9254. Comparing the index evaluation results of different algorithms, the root mean square error (RMSE) predicted by the FWA-BP pattern was about 7.775, and the index effect of MSE was significantly better than the other models. The above shows that the research model has high feasibility and superiority, the advantage of high-speed operation, and provides certain technical means for enterprises to promote great circulation of the international economy.

https://doi.org/10.5573/IEIESPC.2024.13.3.294

(Bhushan Yelure);(Arun Patokar);(Siddheshwar Patil);(Rajesh Mawale);(Sangita Nemade);(Varsha Gaikwad)

Vehicular Ad-hoc networks have gained substantial interest from researchers in recent times. They offer the potential to deploy different applications utilized by intelligent transportation systems, which could improve traffic control and ensure road safety. The vehicle has a significant position within a Vehicular Ad hoc network system. Ensuring security is crucial due to the open wireless environment where data are disseminated. This research paper discusses different types of routing attacks that pertain to data availability and authentication. To assess the impact of these attacks on the routing protocol, a simulation was conducted using the Ad hoc on demand distance vector (AODV) routing protocol with the implementation of these attacks. To perform simulations, trace files for both one-lane and two-lane scenarios were employed. The outcomes of the simulations highlight the influence of an attack on the routing protocol performance, which was compared to the performance of the AODV protocol. In the absence of an attack, the AODV protocol demonstrates an improved collision ratio and packet delivery. In comparison to a Sybil attack, other forms of attacks, such as black-hole, gray-hole, and rushing attacks, have a greater influence on the performance of AODV.

https://doi.org/10.5573/IEIESPC.2024.13.3.303

(Yuandong Li);(Qiong Wang)

Artificial intelligence (AI) has brought great changes to the traditional sports industry. In order to solve the shortcomings of current recognition methods for tennis error training and obtain better recognition effect, we constructed a recognition method based on AI technology. A random projection algorithm was used to reduce the dimension of feature vectors, a CNN (convolutional neural network) model was used to learn training samples after dimension reduction, and a recognition model of wrong tennis training actions was constructed. The preprocessing of data included converting the original Cartesian coordinate system into a cylindrical coordinate system and normalizing the time of a skeletal motion sequence. Experiments show that the recognition accuracy of this model on the NTU-RGB+D dataset can reach 95.34%. The recognition accuracy of this model on the UTD-MHAD dataset can reach 94.12%. Compared with another model, the accuracy of this model was improved, which verified the superiority of this model. It can provide some technical support for the recognition of wrong tennis training actions and improve the tennis teaching effect and students’ learning level.