Hasil untuk "Control engineering systems. Automatic machinery (General)"

Pan Wang, Toru Kurihara

Motion-blurred images are the result of an integration process, where instant light intensity is accumulated over the exposure time. Unfortunately, reversing this process is nontrivial. Firstly, integration destroys the temporal ordering of motion, resulting in ambiguity in the motion direction within a single motion-blurred image. Secondly, unlike conventional single-image deblurring, restoring a sequence of frames can divert the neural network's attention to each individual frame, which results in a decrease in the overall restoration quality of the entire sequence. To address the first problem, we leverage a crucial clue: the correlation image, generated by the three-phase correlation image sensor (3PCIS). This image effectively expresses motion information over the exposure time, which is essential for determining the motion direction of moving objects. We design a two-stream network to restore a sequence of sharp frames from a pair of motion-blurred images and their corresponding correlation images. For the second problem, we propose a temporal progressive learning (TPL) strategy that mitigates the performance degradation caused by network distraction by gradually increasing the number of restored clear frames during training. Experimental results demonstrate that our proposed method surpasses previous state-of-the-art methods on the GoPro public dataset and in real scenes.

Kai Hu, Wenyi Zhang, Bing Qi et al.

The tillage depth (TD) serves as a pivotal criterion for assessing the operational excellence of rotary tillers, yet the current TD control methods are plagued by a myriad of issues including subpar precision and sluggish responsiveness. The present study aimed to develop a high-precision TD monitoring model that incorporates the tilting attitude of the tractor, and investigate the impact of tractor attitude inclination and lifting hydraulic cylinder stroke on TD. The TD stability control system based on electro-hydraulic control was developed, and the model identification method was adopted to derive the accurate control function. The fuzzy adaptive PID (FAPID) method was adopted to effectively improve the response speed and resisting disturbance capacity of the electro-hydraulic system. Then the co-simulation model of the electro-hydraulic control system was constructed. Under the excitation of step and sine functions, the FAPID control algorithm can reduce the rise time by more than 50%, and the displacement tracking error is also effectively reduced. To verify its effectiveness, the experimental platform was constructed, and then field test trials of TD were conducted. The test results indicate that, under various operational conditions, the developed TD control device can effectively reduce the standard deviation of TD by 0.302–0.464 and decrease the variation coefficient of TD by 2.47%–2.92%. The online monitoring and precise control device of TD investigated in this paper can effectively improve the quality of tillage machinery.

Shiping Shao, Abhishek Gupta

We consider risk-sensitive Markov decision processes (MDPs), where the MDP model is influenced by a parameter which takes values in a compact metric space. These situations arise when the underlying dynamics of the system depend on parameters that drifts over time. For example, mass of a vehicle depends on the number of passengers in the vehicle, which may change from one trip to another. Similarly, the energy demand of a building depends on the local weather, which changes every hour of the day. We identify sufficient conditions under which small perturbations in the model parameters lead to small changes in the optimal value function and optimal policy. This is achieved by establishing the continuity of the value function with respect to the parameters. A direct consequence of this result is that an optimal policy under a specific parameter remains near-optimal if the parameter is perturbed slightly. Implications of the results for data-driven decision-making, decision-making with preference uncertainty, and systems with changing noise distributions are discussed.

Yandong Ma, Deli Liu, Kun Ma et al.

Haneul Shin, Daewon Seo, Jihyuk Park

SHI Ke, WANG Yue, ZHANG Zhengfang et al.

Virtual coupling of urban rail train sets represents a cutting-edge frontier in urban rail transit research, with the goal of improving operational efficiency of trains on the existing lines, particularly in the starting and stopping stages. This paper proposes an efficient collaborative control system for virtual coupling. Firstly, a model was constructed to simulate relative braking between urban rail train sets and it was used to generate relative braking distances between trains virtually coupled. Next, an efficient starting and stopping objective function was constructed at the collaborative control layer, and an optimized controller was devised to solve the objective function under constraints, generating key decision-making information for the leading and following trains during the starting and stopping stages. Subsequently, this study focused on the operation sequence planning technology, yielding the executable operation sequences of the leading and following trains, respectively. Finally, an LQR controller was designed, enabling the accurate following and control of operation sequences and the generation of execution instructions. Simulation results showed that the proposed control system for virtual coupling could achieve a parking time interval of less than 1.6s between the leading and following trains in urban rail transit within a speed range of 0~60 km/h, effectively improving the operational efficiency of virtually coupled urban rail train sets during the starting and stopping stages.

Lisen Zhao, Lihua Lu, Xiang Yu

The electricity market is a complex and dynamic environment characterized by a multitude of factors that influence electricity prices. Accurate and reliable electricity price forecasting (EPF) is crucial for market participants, including power generators, consumers, and policymakers. Electricity prices are influenced by temporal dependencies and electricity consumption patterns. Therefore, dependencies across different feature dimensions (cross-dimensional dependencies) and temporal trend information are essential. To address the aforementioned issues, we propose Multimodal Integration and Linear Enhanced Transformer (MILET), which combines cross-dimensional dependencies with single-dimensional modal features. First, we decompose electricity price data into three regular modals using Variational Mode Decomposition and Sample Entropy. This approach enables us to uncover the intrinsic patterns within the variable, thereby simplifying the complexity of the data series. Then integrate these three modals and the original dataset into a five-channel encoder (Modal Integration Encoder, MIE) with both single and multi-dimensional information. MIE is composed of Overall Two-Stage Attention (OTSA) and Long Short-Term Memory (LSTM), where OTSA handles cross-dimensional dependencies, and LSTM addresses long-term dependencies. Additionally, we capture trend information in electricity consumption features through linear layers and linearly integrate the data to obtain the forecasting results. Extensive experimental results on five electricity price datasets demonstrate the effectiveness of MILET compared to state-of-the-art techniques. Our code is available at https://github.com/Lisen-Zhao/MILET/tree/master.

Juan Alejandro Flores-Campos, Adolfo Perrusquía

Elżbieta Doluk, Anna Rudawska, Izabela Miturska-Barańska et al.

The study evaluates the shear strength of the single-lap adhesive joints made of C45 carbon steel. The influence of the overlap length on the shear strength of the adhesive joint was tested. The elongation of the samples was also tested. Before the bonding process, the samples were treated with P180 abrasive paper and degreased. The adhesive joints were made using the Epidian 53/Z1/100:10 adhesive composition. The strength tests were carried out on a Zwick/Roell Z150 testing machine. The maximum value of the shear strength was obtained for the lap Lz1 = 13 mm and the minimum for Lz4 = 19 mm.

Tatyana S. Katermina, Maksim V. Sliva

Heri Andrianto, Suhardi, A. Faizal et al.

An increase in the human population requires an increase in agricultural production. Generally, the most important thing in agriculture that affects the quantity and quality of crops is plant diseases. In general, a farmer knows that his plant is attacked by a disease through direct vision. However, this process is sometimes inaccurate. With the development of machine learning technology, plant disease detection can be done automatically using deep learning. In this study, we report on a deep learning-based rice disease detection system that we have developed, which consists of a machine learning application on a cloud server and an application on a smartphone. The smartphone application functions to capture images of rice plant leaves, send them to the application on the cloud server, and receive classification results in the form of information on the types of plant diseases. The results showed that the smartphone-based rice plant disease detection application functioned well, which was able to detect diseases in rice plants. The performance of the rice plant disease detection system with VGG16 architecture has a train accuracy value of 100% and a test accuracy value of 60%. The test accuracy value can be improved by adding the number of datasets and increasing the quality of the dataset. It is hoped that with this system, rice plant disease control can be carried out appropriately so that yields will be maximized.

Keisuke Shimako, Masanobu Koga

In this paper, an algorithm for index reduction of differential algebraic equations (DAE) is proposed. Pantelides algorithm has already been proposed as an algorithm for this purpose. This conventional algorithm has succeeded in reducing the calculation time required for index reduction. However, there exist some DAE systems whose index cannot be reduced correctly with the conventional algorithm because it uses sufficient condition to reduce the index. We propose an algorithm to modify the solution obtained by Pantelides algorithm to deal with a wider class of DAE systems. The proposed algorithm deals with DAE systems by using the necessary and sufficient condition to reduce the index systematically that cannot be reduced by the conventional algorithm. We implemented the proposed algorithm as the functions on Maxima and evaluated the algorithm by using some examples.

J. D. Gil, L. Roca, M. Berenguel

Membrane distillation is a termally-driven separation process under investigation. This technology stands out for the simplicity of the process and for its low operating temperature, which allows it to be combined with low grade solar energy. Thus, membrane distillation has become a promising, effcient and sustainable solution for the development of small-medium stand-alone desalination facilities to be implemented in offgrids areas with good irradiance conditions. However, in order to develop this technology on an industrial scale, research must continue to improve aspects related to both the design of membranes and modules and their operation. Regarding the operation, the development of models and control techniques play a fundamental role. This paper presents a review of the control and modeling techniques applied in this field, describing the main methodologies employed and the future challenges to be addressed, also including an illustrative example.

Carlos Balaguer

Prangya Mohanty, Rabindra Kumar Sahu, Sidhartha Panda

A hybrid Many Optimizing Liaisons Gravitational Search Algorithm (hMOL-GSA)-based fuzzy PID controller is proposed in this work for Automatic Generation Control problem. MOL is a simplified version of particle swarm optimization which ignores the particle best position consequently simplifying the algorithm. The proposed method is employed to tune the fuzzy PID parameters. The outcomes are equated with some newly proposed methods like Artificial Bee Colony (ABC)-based PID for the identical test systems to validate the supremacy of GSA and proposed hMOL-GSA techniques. Further, the design task has been carried out in a three-area test system and the outcomes are equated with newly proposed Firefly Algorithm (FA) optimized PID and Teaching Learning-Based Optimization (TLBO) tuned PIDD controller for the identical system. Better system response has been observed with proposed hMOL-GSA method. Finally, sensitivity study is being carried out and robustness of the proposed method is established.

Shaojiang Zhang, Yanning Guo, Qi Zhu et al.

Localization is a critical issue in autonomous navigation and path planning. The traditional localization method for automatic driving is to use the GPS. But in many cases, such as near the high buildings, under the viaduct, in the basement or the tunnel, GPS signal will disappeared or be weakened, resulting in localization failure. This paper proposed a precise localization method based on Lidar, IMU and wheel odometer combined with point cloud data map matching. In this paper, information in the form of depth point cloud collected by Lidar is matched with the pre-known point cloud map data, and a Point-to-Plane Iterative Closest Point algorithm (PP-ICP) is introduced. In order to avoid the mismatch and localization failure, the start of the autonomous vehicle is set to the initial coordinates of the point cloud map data. Then, the predicted state equation which consists of the throttle control and steering wheel control of the autonomous vehicle is established. The first observation equation consists of the acceleration and angular velocity measured by the IMU. The second observation equation consists of the position and attitude measured by the Lidar matching point cloud, along with speed and distance measured by the wheel odometer. Finally, the extended Kalman filter is used to fuse the information data of the three sensors and thus update and correct the localization. Experiments using multi-sensor fusion were carried out in underground garages and the experimental results showed that the robustness and positioning accuracy can meet the engineering requirements.

Aitziber Iglesias, Markel Iglesias-Urkia, Beatriz López-Davalillo et al.

Internet of Things (IoT) devices are usually advanced embedded systems that require functionalities monitoring and control. The design, development and validation of these devices is complex, even more when communication capabilities need to be included. In industrial environments, where safety is of critical importance, reducing this complexity can help to achieve the vision of Industry 4.0 by reducing development time and costs as well as increasing quality. To this end, the use of Model-Driven Engineering (MDE) methodology and the Software Product Line (SPL) paradigm is becoming increasingly important as they help to accelerate and ease the development of software, while reducing bugs and errors. Thus, in this work we present TRILATERAL, a SPL Model Based tool that uses a Domain Specific Language (DSL) to allow users to graphically define the IEC 61850 information model of the Industrial Cyber-Physical System (ICPS). TRILATERAL automatically generates the source code for communicating devices with the monitoring framework, also supporting a variety of communication protocols, namely HTTP-REST, WS-SOAP and CoAP in order to control/monitor any ICPS. In addition, the solution was evaluated deploying it in different industrial domains (Wind Farm, Smart Elevator, Catenary-free Tram) from which we gained important lessons.

Halima Atitallah, Asma Aribi, Mohamed Aoun

Hidetaka Taniguchi, Hiroshi Sato, Tomohiro Shirakawa

The neural network is one of the most successful machine learning models. However, the neural network often requires large amounts of well-balanced training data to ensure prediction accuracy. Meanwhile, human learners can generalize a new concept from even a small quantity of biased examples, simultaneously enlarging knowledge with an increase in experience. As a possible key factor in the ability to generalize, human beings have cognitive biases that effectively support concept acquisition. In this study, to narrow the gap between human and machine learning, we have implemented human cognitive biases into a neural network in an attempt to imitate human learning to enhance performance. Our model, named loosely symmetric neural network, has shown superior performance in a breast cancer classification task in comparison with other representative machine learning methods.

KN Dewangan, S Rakheja, P Marcotte

The gender and anthropometric effects on vibration absorbed power characteristics of the seated body are investigated through measurements with 31 males and 27 females considering two different back support conditions, and three levels of vertical vibration (0.25, 0.50, and 0.75 m/s 2 rms acceleration) in the 0.5–20 Hz frequency range. The absorbed power responses for the males and females revealed strong gender effect, which could be mostly related to differences in body mass of the two groups. Subsequent analyses were conducted considering different datasets grouped corresponding to three ranges of the body mass-, build-, and stature-related parameters for both the males and females. Notable differences were evident in the absorbed power responses of the males and females with comparable anthropometric dimensions. Males revealed significantly higher peak and total absorbed power responses compared to the females of comparable anthropometric dimensions, except for the lean body mass. The differences, however, were relatively small in the data for males and females of comparable body mass. The peak power for the females, invariably, occurred at a lower frequency than that for the males. The total absorbed power responses revealed some degree of correlations with the body mass, lean body mass, body fat, and hip circumference ( r 2 >0.60), irrespective of the back support condition and excitation magnitude for both the genders.