Search

Microgrids are one example of a low voltage distributed generation pattern that can cover a variety of energy, such as conventional generators and renewable energy. Economic dispatch (ED) is an important function and a key of a power system operation in microgrids. There are several procedures to find the optimum generation. The first step is to find every feasible state (FS) for thermal generator ED. The second step is to find optimum generation based on FS using incremental particle swarm optimization (IPSO), FS is assumed that all units are activated. The third step is to train the input and output of the IPSO into deep learning (DL). And the last step is to compare DL output with IPSO. The microgrids system in this paper considered 10 thermal units and a wind plant with power generation based on probabilistic data. IPSO shows good results by being capable to generate a total generation as the load requirement every hour for 24 h. However, IPSO has a weakness in execution times, from 10 experiments the average IPSO process takes 30 min. DL based on IPSO can make the execution time of its ED function faster with an 11 input and 10 output architecture. From the same experiments with IPSO, DL can produce the same output as IPSO but with a faster execution time. From the total cost side, wind energy is affecting to reduce total cost until USD 22.86 million from IPSO and USD 22.89 million from DL.

This study aimed to propose a control scheme to optimize the active power load using the pulse-width adjustment technique and the MOSFET driver. The power used and stored in the battery is an input for the fuzzy system whose information is obtained through the current (ACS 712) and voltage sensor readings. The use of fuzzy to control the power of two 5-watt lamps is more efficient than the manual technique using an ordinary switch. This is because fuzzy only consumes 3.25 Watt/hour while the manual technique requires 5.7 Watt/hour. Based on the linear regression-based estimation, the fuzzy technique lasts ±17 hours from the initial power of 55.82 Wh, or 5.5 hours longer than the manual technique that lasts only ±11.5 hours from the initial power of 65.62 Wh. Therefore, this study adjusted the load power to extend battery life and increase solar energy use efficiency and innovations in load control based on available resources.

Indonesia has an electrification ratio target of 100% by 2020. This target is supported by the projected electricity demand to increase more than seven times by 2050. Therefore, the government needs to build new generators to increase the national electricity production capacity. The Ngadingoro River in Trenggalek Regency, with its fast-flowing water, is an area that has the potential to become a power plant. This study aims to create a portable pico hydropower plant to optimize the use of renewable energy to generate electricity. The power plant made in this study utilizes the power of river water as a driving force. It modifies the floating ponton's hull to become portable in adjusting the water level and optimize the kinetic energy of the water flow. The results showed that this portable floating hydropower works well by producing 720 watts of electricity/day.

It is with great pleasure that I introduce the proceedings of the 3rd International Conference on Advance Engineering and Technology (ICATECH 2021) is an international seminar organized by Institut Teknologi Adhi Tama Surabaya (ITATS). ITATS is private university located in Surabaya, East Java, Indonesia. The conference is held on October 2nd 2021. This seminar has the main purpose to bring researcher and scholar to share their knowledge and experience in Computational Science, Environmental and Earth science, Applied Science and Technology, Chemical physics and physical chemistry, Renewable Energy and Integration Electrical Science Material Physics, materials analysis, and characterization. The conference serves as an excellent opportunity to meet each other and to exchange ideas with theme of “Opportunities and Challenges for Research and Technology Enhancement in Post Covid-19 era”. List of Editor, Special Thanks To Our Keynote Speaker, Organizing Committee, Scientific Commiittee are available in this pdf.

Based on BPJS Employment data (2018), 157,313 accidents case happened in Indonesia. Lack of awareness and discipline of workers in the use of Personal Protective Equipment (PPE) is a main factor of the occurrence of work accidents. This identification system uses image processing which is modified by the Convolutional Neural Network (CNN) method, where this algorithm will process and analyze images of workers using PPE. The APD detected in this study is on head area such as Safety Helmet, Safety Glasses, Safety Masks, and Safety Earmuff. Twelve classification datasets were prepared for the training process with a total number of 917 image datasets. The input of this study is an image capture of workers using PPE and additional Inductive Proximity sensor is used to detect Safety Shoes. The output of this study is the results of the classification of PPE completeness which are used by the workers, with green 12 Volt DC lamp indicator for complete category indicators and red 12 Volt DC lamp indicator for the indicator if any or all of the PPE is not used. Based on the test result of this study, it obtained that the percentage of accuracy when tested real data time was 85.83 % for respondents who were included in the dataset, 80 % with percentage for respondents who were not included in the data set, and 73.34 % with percentage for female respondent.

The number of water accidents in Indonesia has increased significantly in the last five years. The Evacuation Team must work quickly to carry out a rescue mission for the victims who nearly drowned. Therefore, this research made Aufloat (Autonomous Float) based on LORA (Long Range) and Artificial Intelligent using image processing with Haar Cascade method as an aid for drowning victims because of water accidents. With this feature, it aims to detect victims and move quickly to victims in evacuation activities. From the results of trial, Aufloat, weighing ± 5 kg and payload 96 kg, has an average speed of 0.94 m/s from 20 trials at different distances, the camera on Aufloat can detect victims well at a maximum distance of 3 to 4 meters and can function well at transmission distances of up to 700 meters.

Prosthetic legs are a tool that is often used by a disabled person who has not legs. It is quite easy to use prosthetic legs for amputated legs in the calf but not with those with disabilities who have amputated their legs in the thigh because they have lost leg joints. There are already a handful of people making conventional prosthetic legs, but because the motion of the knee joint on conventional prosthetic legs only uses a work system of mechanical motion, it is still difficult to use for daily activities, especially in carrying out activities up and down stairs. Then there are also prosthetic legs that use EMG (Electromyography) muscle sensors, but the use of EMG sensors has disadvantages in making them require a lot of adjustments and research for different users, because the feet have the same function namely to walk, so that the movement every human being has the same movement. With this fact, in this journal discussing the development of fake feet using gyro electric sensors, the authors have the idea of using gyro, axelero and heavy sensor sensors in the fabrication of prosthetic feet using the fuzzy logic method. Gyro, axelero and heavy sensor sensors are installed on the lower leg or calf, so that in making prosthetic legs on new uses there is no need to make complicated adjustments. In applying this fuzzy logic method, modeling is made that can simulate a person's movements while walking where the user only needs to swing this fake leg like someone who has whole / perfect legs.

the purpose of this study is helping persons with disabilities for moving a wheelchair using brain waves, the wheelchair will move depending on the watched color by user. There are three colors to be detected by the system; they were red, green and blue. This tool is called the BWD (Brain Wave for Disabilities). Brain waves were received with the neurosky Mind wave Headset to produce electroencephalograph (EEG) waves. EEG waves go through the extraction process using the FFT algorithm according to the rules of nyquist frequency. Each color was taken as many as 50 times for this experimental case. Each retrieved data in one color has 500 data points. Then the data points are used as input to the classification method called Deep Learning algorithm. The results of the classification are used to move the DC motor in a wheelchair. The wheelchair will move according to the color seen by the subject. The success rate in this research was 66, 67%.

Industrial Revolution 4.0 is the center of automatic technology development and adoption. This applies to the development of industrial automation and information technology. In the manufacturing industry, machines have been able to work autonomously to carry out the production process quickly and precisely. Even in the development of information technology, expert systems have been embedded in various smart phones. There are many things that can be controlled through smart phones that are connected to the internet network. Not only in manufacturing industries or offices, the development of industrial technology 4.0 has also begun to be implemented in homes. Automatic robotic cleaning technology, or smart home applications can be used commercially. In fact, using people tracking technology, automatic trolleys have been applied to help consumers in supermarkets. In this paper, person-following robot was developed. We use depth cameras to recognize human movement. 3D data is used as a reference value in a human follower system. Minimizing computing time, the dynamic decision tree method is used. This offers lighter and faster computational processing than using the fuzzy or NN method. Based on the testing result, a good robot performance is obtained. Robots can follow human movements in real-time on various testing paths.

in this research, a tool that can provide information about object around is made. This tool can also estimate distance of detected object through camera which is combined with glasses, to ease blind people who use it. This tool is certainly can help them to identify object around and improve their skill and ability. This tool use camera as main sensor, which works like human eyes, to provide real time video as visual data. The RGB visual data is processed using Convolutional Neural Network which has 176x132 pixels by convoluting 2 times. It produces smaller pixels with size 41x33 pixels, so weights is obtained for classification using back propagation and determined dataset. After getting detection result, the next step is a find centroid value as center point for measuring the distance between objects and cameras with Stereo Vision The results is converted into sound form and connected to earphones, so blind people can hear the information. The test results show that this tool can detect predetermined objects, namely humans, tables, chairs, cars, bicycles and motorbikes with an average accuracy of 93.33%. For measurements of distances between 50 cm to 300 cm it has an error of around 6.1%

The latest in unmanned aerial vehicles (UAVs) and associated sensing systems make these increasingly attractive platforms to the remote sensing community. A large number of spatial details contained in these images opens the door for advanced monitoring applications. In this paper, we use this cost-effective and attractive technology for the automatic detection of green open spaces. Given a UAV image of trees acquired, then, we analyze these Convolutional Neural Networks (CNN) points of the prior classifier trained on a set of trees and no trees points. As output, CNN will mark each detected tree by super pixel. Then, in order to capture the shape of each tree, we propose to merge this pixel-level segmentation with a method based active contour on the Color threshold. Finally, we further analyze the texture of regions with pixel-level segmentation and use summing pixel to distinguish trees from other vegetation. Experimental results obtained in UAV images from extensive calculations using the program that has been made and the existing provisions get a result of error of 7.256% on the first trial, the second experiment is 5.156%, and the third experiment is 3.126%.

The face is one of the easiest way to identify individual and to distinguish. Therefore, the face recognition system is usually needed in the security system in the restricted rooms of the company. This research is to minimize all fraudulent actions such as theft of the company data. In this final project, the method used is histogram equalization and fisherface. The main step in this security system is that the user's face will be taken using a webcam. Then the process of face recognition uses the histogram equalization and the fisherface method using a PC (Personal Computer). Fisherface is a combination of Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) methods. When an RFID sensor matches the employee data, the camera will capture and the process of face recognition will be run. After that the face data will be matched with the existing face data. When the both is match, the PC sends a command to the Arduino microcontroller to open the solenoid door lock. So that the security systems through face recognition can be more effective than conventional security systems. The test result of the face recognition system which has been done in this Final Project, has a success rate of 88.33% which was obtained from 120 times of experiments consisted of 12 poses. The test success level of the security system with 3 correspondents was 88.33%.

One indicator of road’s feasibility can be seen from the conditions on the road surface. Road damage that often occurs is the number of potholes or holes [1]. In this study, we made an inspection device that can detect hollow roads automatically. Where the results of the detection will be sent to the server along with the coordinates of the location of the hole that will be processed and used as data on information on roadworthiness in an area. So when conducting road inspections we do not calculate the number of manually perforated roads so that the inspection process will run faster and more efficiently. This tool works with the camera as the main sensor, the camera works with the eye to provide visual data in the form of real-time video [2]. The visual data is processed using the Convolutional Neural Network algorithm which has 480 x 320 pixels with convoluted 2 times and produces smaller pixels, so we obtain weights to be classified with datasets [3]. These results will result in a decision whether there is a hole or not on the road that has been inspected by this tool. In addition to knowing where the location of the damage is this tool is equipped with GPS. If a hole is detected, the tool will take a picture, then send the image and coordinates to the server, so that the server can find information on the hole in an area [4]. This study obtained a success percentage of 92.8% in detecting holes using the CNN (Convolutional Neural Network) method.

In this study, attendance system was created by combines RFID attendance technology and face recognition. The attendance system is an important thing to facilitate attendance data retrieval. Attendance system often encountered in the technological era such as today is the RFID attendance system. Although the attendance system is already sophisticated, this system still has shortcomings such as the occurrence of entrusted cards. In this research, the writer has two inputs, namely ID on the RFID card and a face image. Face images will be process by two methods there are the Viola-Jones method that use to detecting the face objects in the image, and the Local Binary Pattern Histogram method as face recognition. Furthermore, if the face was recognized then the RFID card ID data and face recognition data will compare, the results of this process will be attended data in the database. Based on the results of test have been done by the author, the response system shows very good results. In the first test of the RFID sensor in 11 tests, the sensor was able to distinguish all cards that had been registered and cards that had not been registered. This test is carried out from distance of 1 cm to 4 cm. Furthermore, in the second test of facial recognition, in 100 trials of facial recognition testing with different facial poses, the system can recognize faces 93 times so system has 93% success rate. This success rate is influenced by several factors such as the test carried out at 364 lux illumination strength, the test distance at $\gt $ 20cm and $\lt $70cm, as well as the maximum testing angle 45°.

Wind energy is one of the several types of renewable energy that exist today. however, wind energy has a high degree of uncertainty due to weather effects. Wind speed prediction is needed to determine the energy that wind turbines can produce at each unit. For optimizing wind speed schedulling, the accuracy of wind speed prediction is considered. Extreme learning machine (ELM) and neural network (NN) is implemented to predict hourly wind speed for 24 hour and power generation from wind turbines can produce. Wind speed probability data is taken from sidrap wind farms in indonesia. To determine the performance of wind predictions based on the error value between actual and predicted, mean absolute percentage error (MAPE) is applied.

The development of installed capacity in power plants in 2018 was 41.696 MW or increased from the previous year in 2017 amounted to 39.652 MW (PLN Statistics, 2018). This has an impact on the reduced availability of fuel due to overexploitation. The highest energy sold per customer group in 2018 was the household customer group that was 41.7% higher than the industrial sector customer group by 32.8% (PLN Statistics, 2018). At present the use of electronic equipment for household needs is increasingly diverse. Many equipment that is often used in daily life is electronic equipment that is inductive load. Inductive load causes the value of the power factor to fall so that the power usage (Watt) becomes less than optimal. To overcome the problem caused by the large number of inductive loads a reactive power compensator is needed which is to use a capacitor. In this final project, a system designed to be able to measure and correct power factors automatically uses the Neural Network method and can monitor power online based on IoT. The results of testing the power factor improvement system were 97.8% successful in the trained electric load and 94.8% in the untrained electrical load.

The application of multivariate statistical techniques including cluster analysis and principal component analysis-multiple linear regression (PCA-MLR) was successfully used to classify the river pollution level in Taiwan and identify possible pollution sources. Water quality and heavy metal monitoring data from the Taiwan Environmental Protection Administration (EPA) was evaluated for 14 major rivers in four regions of Taiwan with the Erren River classified as the most polluted river in the country. Biochemical oxygen demand (6.1 &plusmn, 2.38), ammonia (3.48 &plusmn, 3.23), and total phosphate (0.65 &plusmn, 0.38) mg/L concentration in this river was the highest of the 14 rivers evaluated. In addition, heavy metal levels in the following rivers exceeded the Taiwan EPA standard limit (lead: 0.01, copper: 0.03, and manganese: 0.03) mg/L concentration: lead-in the Dongshan (0.02 &plusmn, 0.09), Jhuoshuei (0.03 &plusmn, 0.03), and Xinhuwei Rivers (0.02 &plusmn, 0.02) mg/L, copper: in the Dahan (0.036 &plusmn, 0.097), Laojie (0.06 &plusmn, 1.77), and Erren Rivers are (0.05 &plusmn, 0.158) mg/L, manganese: in all rivers. A total 72% of the water pollution in the Erren River was estimated to originate from industrial sources, 16% from domestic black water, and 12% from natural sources and runoff from other tributaries. Our research demonstrated that applying PCA-MLR and cluster analysis on long-term monitoring water quality would provide integrated information for river water pollution management and future policy making.

This paper proposes a method for constructing a harmonic analyzer prototype used to monitor harmonics caused by a power source generated by a car inverter. The prototype is called Harmonic Monitoring of Car's Inverter (HMC'sI). hMc'sI consists of a current sensor and a microprocessor for data processing and monitoring HMC'sI utilizes PC with Matlab software. The advantage of HMC'sI is that the harmonics that occur can be directly monitored on matlab. so that the analysis of the harmonics that occur as well as simulation steps to reduce the harmonics can be directly simulated. In addition, the price of HMC'sI is very cheap, about one- twentieth times the price of harmonic analyzer commonly sold in the market. The observation is focused on Harmonic 3rd, 5th, 7th, and 9th, then calculated Total Harmonic Distortion (THD) by reference to the standard IEEE 519. from the experimental results it is known that HMC'sI can detect harmonics that occur well with a standard deviation of 1.5% compared to laboratory scale Harmonic Analyzer.

This paper proposed new technique for improving smart meter. This technique is built based on Non-Intrusive Load Model (NILM) which is combined with time series modify (lag-1). This technique is employing modification of time series data to predict the operating of electrical appliances which is operated simultaneously. The advantage of the technique is capable to identify using of energy consumption of appliances without adding sensor in each appliances. Another advantage of this method is the simplification of the Neural Network (NN) training process, because with the concept of lag-1, each appliance requires only one data record. Signals from current sensors are processed by the microprocessor to identify the type of appliance currently operating by using NN. Data resulted by NN is sent to the display and also sent to the SD Card which can show the bill of each electrical equipment in detail. From the experiment result, it can be proof that smart meter capable to identify the use of appliances and also capable to monitor the use of energy consumption real time with 5% error tolerance in averages. With this performance, the smart meter has big chance to implement in the real systems and mass production.

This paper presents a comparison of two methods of artificial intelligence which applied in Wallfollowing Autonomous Mobile Robot; both of them are Neural Network Backpropagationand Fuzzy Logic. The robot has three input variables and two output variables. The inputs are distance between the robot and the wall which is sensed by HC-SR04 ultrasonic sensors. The output variables are the speed of the two wheels which is driving by 12 Volt DC motor. In this case mobile robot is designed to avoid the collision with any obstacles like wall or other mobile robots. In this implementation mobile robot is designed with a numbers of ultrasonic sensors and placed on certain position like center front, left front and left back. The sensor will send the data in real time. After being processed, the input produces output in form of speed value governing motor rotation mounted on both wheels of the robot to find the optimum point. In this comparison, both methods Backpropagation Neural Network and Fuzzy Logic are treated the same. Wallfollowing Autonomous Mobile Robot is using Atmega2560 microcontroller. The logic is uploaded to the microcontroller. The result of the comparison of these two methods when applied in Wall-following Autonomous Mobile Robot is the movement of the robot using Neural NetworkBackpropagation is faster than using Fuzzy Logic Controller.

In today's era the technology development is growing so fast and change the system that was originally manual into a system that all operate completely automated. By automating a production system, the number of product will increase. One of the civil materials industry that still use a lot of manual system is light brick making industry or commonly called as CLC (Cellular Lightweight Concrete) brick. A microcontroller-based light brick manufacturing system needs to be applied to optimize the production. The use of arduino mega 2560 microcontroller as the main control in the mixing automation system on this light brick making device will make it easy to do. This tool measures the weight of sand and cement using HX711 load cell sensors as a mass sensor and flow meter sensor as the required water flow sensor in the mixing process. As an actuator the Selenoid valve is useful as an automatic valve on the hole tube of water and foam, while selenoid door lock is useful as an automatic valve on sand holes, cement and stirring tube. The motor is controlled by H- Bridge Driver Motor IBT arduino. To make the system easier (Human Machine Interface) Visual Studio.Net software is applied to monitor and control the process in CLC brick making. With the prototype of automatic mixing system, the used of microcontroller in light brick making is expected to improve the industrial sectors or light brick manufacture who want to optimize the quality and quantity of its production. Automation on this system also makes all processes run fast and does not require a lot of manpower.

This paper has proposed a prototype of a control system in deaerator storage tank level at Industrial Steam Power Plant based on artificial intelligence (AI). There are two kind methods of AI which are implemented in this research, first is Back Propagation Neural Network (BP-NN) and the second is Extreme Learning Machine (ELM). The proposed method is aimed to improve the performance of an existing Proportional Integral (PI) control method. The input variables are error level and load condition. The output variables are control valve percentage and indicator value. From the experiment, the result proved that ELM is fast superior to BP-NN according to the time of training process and error tolerance. Prototype-based on ELM is also working properly with an error tolerance of 0.15 %.

The world's economic instability makes people very sensitive to the costs incurred to consume electrical energy. In this paper proposed smart meter that can record the consumption of electrical energy of any electrical equipment. The proposed method is employing Non-Intrusive Load Monitoring (NILM) concept which is combined with time series modify data processing. The advantages of the proposed method are the efficiency of the current signal reader and the least amount of data taken in the training process of artificial neural network — Extreme Learning Machine (ELM). The proposed method was using transient signals and steady state signals as sign to identify the condition of equipment ON or OFF. The time series modify method is helpful for data retrieval when many electrical devices are operated. From the experiment results, smart-meter are expected to be utilized to make an electric bill with details of the load usage of any electrical equipment.

The proposed paper is an auto adjustment of instant noodle dough thickness on a roll press machine. In this plan will be implemented two artificial intelligent methods of Neural Network Back propagation (BP-NN) and Extreme Learning Machine (ELM). The method is intended to improve the performance of the controls on this machine. The input variables are the height and thickness of the dough by using the sharp sensor. While the output plan is two, dc motor and stepper motor. From the results of experiments that have been done, it is concluded that the use of extreme learning machine method is better than the Neural Network method. This is evidenced from the comparison of the average error output value on the ELM of 0.002440673 for training data and 0.6337302 for validation data. While the average error of output value on NN is 0.006638958 for training data and 0.568553072 for validation data. The training time of ELM method is 0.0072812 second and NN method is 0.501274 sec.

On this paper, the predicted results of the two types of methods will compare. This is very important because, the error in the determination method will cause the prediction result is not optimal. Two types of methods to be coordinated are the Extreme Learning Machine and Back propagation Neural Network. Testing will be done by entering the calculation from each method to Hand Typist Robot to perform the same task. This robot is used to help Quadriplegic Person in operating a computer keyboard and has 4 Degrees of Freedom (DOF) in both arms. The actuator for all DOFs is Dynamixel AX-12A. This robot processed CMPS10 sensor data which sense the changes provided by Quadriplegic Person. The change of motion will be processed using a method to predict the movement of each Dynamixel AX-12A motor. The most optimal method will produce the lowest Mean Square Error (MSE).

This paper presents design and experimental studies of Extreme Learning Machine (ELM) to control temperature and humidity of oyster mushroom farm house. The ideal temperature to optimize the growth of oyster mushroom in low lying areas is for about 28° Celsius and 80% of humidity, while the current method for controlling temperature and humidity is done by conventional manner using manual sprayer. Given these problems, a Single Layer Feed Forward Neural Network (SLFN's) with modification of H inverse matrix versus target matrix or also known as ELM can control the temperature and humidity of oyster mushroom farm house more faster and effectively than previous research. DHT11 sensor is used to read the temperature and humidity value. Exhaust fan and mist maker are used for conditioning the control variables. Several beginning conditions were built to compare ELM with previous methods such back propagation neural network and zero order FLC in term to find the suitable methodfor this problem.

This paper is presenting a design and research studies in industrial steam power plant system called “control and monitoring system to optimize the combustion process in the furnace boiler prototype with comparison of neural network and extreme learning machine”. Comparison between Neural Network and ELM (Extreme Learning Machine) methods will be used to this combustion control system and will be implemented in a prototype with microcontroller. This prototype is using the value of temperature sensor and value of smoke sensor in the furnace as parameter of heat to control the flow of air and fuel oil. The temperature sensor in this research is type K Thermocouple. The smoke sensor is MQ sensor. This prototype also used fan and pump oil as an actuator. Fans are used to supply the oxygen and pump is used to supply the fuel oil. From the experimental result, this prototype shows the optimization of combustion system using ELM (Extreme Learning Machine) method can work well compared with NN (Neural Network) method. ELM Control System has a very good response and it can work well (RMSE = 6,32456E-05). So, if the system is applied in the industrial steam power plant, it can improve the performance of combustion control systems and able to save the fuel.

Renewable Energy became a popular topic at the beginning of the third millennium in order to seek an alternative energy and it is expected to be environmental friendly. This is caused by the consumed daily energy like coal, petroleum, natural gas and others are no longer exist and cannot be renewed. Wind energy is one of the energy which is easy to use and can be obtained through wind energy conversion system or it is usually called by wind turbine. A wind turbine should have an equal or close power coefficient (CP) value to the maximum wind turbine standard value. CP is the determining factor for the advisability of wind turbines, because the greater value of CP the wind energy conversion will also be greater. IPSO is a combination of Particle Swarm Optimize (PSO) technique with Incremental Social Learning (ISL). The addition of the ISL algorithm allows this method to obtain global optimum value faster because ISL is a method of adding particle at specific time based on existing information. In this research PSO and IPSO will be compared in order to find an optimum CP from wind turbine prototype. Comparison result with mathematical approach is produced MSE = 0.0258for PSO and 0.0222for IPSO.

One Chemical Company is a state-owned factory engaged in the production of fertilizers. Especially in factory which is a supporter of air needs for line to line needs in Plant Air Receiver. The air requirement is supplied by three air compressors using an electric power. Air is accommodated on a large tube that is called D923 Plant Air Receiver tube. After meet the required air pressure, air will be channeled to the Plant Air Receiver line. However, three compressors still use manual systems where the operator goes down to the plant to turn off and turn on the compressor and the compressor is still running independently. If the compressor burns continuously in hot conditions then the compressor lifetime will shorten. With the above problems, the researcher tries to create a system of air pressure plant automation receiver by using microcontroller, as a main control of the system are using ELM and NN. Both of them are used for controlling the air valve filling and air pressure measurement process using MPX5500DP sensor and compressor temperature measurement using DS18B20 to make two compressors as main compressor and backup compressor. It is expected to be used to assist in the process of control and monitoring of air filling so it can be controlled through control room.

This paper presents design and experimental studies of vending machine for office stationery Transactions. The advantage of the proposed vending machine i.e. Transaction can be done by using short message system (SMS), all transaction can be monitored online by owner by using Android, the vending machine has feature early warning system (EWS) when system in trouble, and it also equipped with battery backup when electricity cut off, No need to make special agreement with bank or telecommunication provider. The Smart Vending Machine is built by using common hardware component such as Arduino as controller, Wavecome as SMS Gateway module, Servos, Power Supply, Battery as power back up, Keypad and button as input, LCD 16×2 as Display. From the several test including normal transaction, online monitoring, and early warning system for electricity supply. The Smart Vending Machine was successful. And it has a big possibility to be mass production.

This paper will present an implementation of Extreme Learning Machine (ELM) in Prototype of Hand Typist Robot (HTR). HTR is Typist Robot which is designed for quadriplegic people. HTR consists of two robotic arms with three dynamixel AX-12 that mounted on each arm. It is mean that each arm has 3 DOF. To operate HTR, user has to equipped with compass sensor (CMPS10), installed on the part of body that has good function. In this paper ELM is used to map and make decision between the signal which sending by CMPS10 and position of alphabet that will be reached by Robot Arm. The advantage of ELM is superior in training process and easy to implement. Using ELM, the relationship between input and output can be present only using one simple matrix. From the experiment result shown that 73 keys of computer keyboard can be reached by HTR with an error 5%. The error is accumulated errors which is caused by vibration of dynamixel AX-12 when it is moving. To minimize the error the HTR need to reset regularly.

This research has made a tracking system to track the target, the purpose of this prototype is being able to be applied in the warship cannon guns. The system principal is tracking the targets that have been chosen to be targeted by the operator. This device works with the camera as the main sensor, the camera works to provide visual data, the visual data as an object was selected and processed to be targeted. The visual data is processed using the Optical Flow Algorithm (OFA) method by calculating the brightness intensity of the reference. The next step is finding the centroid value as the centre point for distance measurement. These results were converted into a microcontroller program that will move the direction of the tracker to follow the target which captured by the camera. The results of the object tracking test on this tool from 10 attempts can detect objects with an average accuracy of 100 percent at a distance of 1 to 9 meters. For testing the target lock, this system can reach the set point value within 2 seconds with the steady state error of 0.625 percent and the selection of the best proportional constant of 0.2.

A voltage source converter (VSC) is one of the most widely used power converters in a power system. In this paper, a time-domain-based accelerated steady-state method is proposed to solve for a closed-loop pulse-width modulated (PWM) VSC with dq-frame controllers, which is able to account for the harmonic interactions between the converter and the rest of the power network, between the AC and DC sides of a VSC, and between the converter and its controllers. The proposed time-domain method is based on the modified time-domain shooting method, where the Jacobian matrix is updated by the quasi-Newtons method. This will drastically increase the computation efficiency as it avoids re-evaluating and inverting the Jacobian matrix, whose size is usually very large for a PWM-VSC due to high number of times of switching. All the results are shown to be consistent with those obtained by a PSCAD/EMTDC model, which has been validated with the experiment in a previous publication. © 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

The productivity of oyster mushroom cultivation in low-lying areas are still not optimal. This is due to the cultivation of oyster mushrooms needs ideal temperature and humidity (temperature 22–28 ° C with a humidity of 60% – 80%), while nowadays temperature and humidity preservation process is done in a conventional manner. Given these problems, the researchers gave the solution by creating a tool that able to work automatically to monitor and control the temperature and humidity in oyster mushroom cultivation problem based on microcontroller. Inputs used in these system are the value of temperature and humidity data readings from DHT11. While the output of the system is two actuators, the first is the exhaust fan and the second is mist maker. In the operation of the appliance automatically there will be two choices of data processing methods are applied, the method of Fuzzy Logic Control (FLC) and Feed Forward Backpropagation Neural Network (BPNN). Performance tools based on the application of these two methods will be compared to determine the most optimal and effective method when it applied to the tool to automatically control temperature and humidity oyster mushroom farm house. Based on the test results and data analysis, the tools can work well and also perform that optimal and effective data processing method is Neural Network with an average conditioning response time of 69.8 seconds to reach the ideal temperature and 113.4 seconds for the ideal humidity.

Nowadays, the research in robotics field is growing. One of the studies in robotics is the control method of the robotic arm movement. In this research, a 3 DOF arm drawing robot was built. An inverse kinematic models of the robot arm is made using artificial neural network method. Artificial neural network model was implemented in a GUI application. The ANN model can work in real-time to control arm robot movement to reach certain coordinates. Based on test results, the inverse kinematic models of the arm drawing robot had an error rate under 2%. It is of 0.16% for X coordinate and 0.46% for Y coordinate.

Finding out about the damage of the induction motor particularly in the rotor bar is essential to ensure the continuity of the manufacturing process and to reduce the high cost of breakdown maintenance. In the past decades, the difficulty level to find fault in rotor bars has attracted many researcher to develop the solving method. Therefore, this work presents an alternative to discover fault on rotor bars based on analysis of stator currents (MCSA) using Discrete Fourier Transformation (DFT) and Extreme Learning Machine (ELM). In this study, ELM applied to detect abnormalities and fault of rotor bars on induction motor with no load condition. The performance of proposed algorithm would compared with Constructive Back Propagation Neural Network (CBP-NN). The structure of both methods consist of a single hidden layer with 20 neurons that activated by a tangent sigmoid function. In the experiment, the data input for all algorithms came from the normalized of DFT outcome. The test result shows that ELM is faster compared with CBP-NN about 0.0467 seconds in terms of time training. Although the accuracy of their training has a wide error deviation but for the new data pattern recognizing or its variation, ELM provides the output prediction closer to the target validation than CBP-NN. Thus ELM can be used as a top priority because of the high level of time efficiency and accurate predictions.

This paper present a new robust microgrid power flow based on Sequential Power Flow (SPF) method, which is capable to account for a DG unit. The SPF method can readily accommodate for PV buses as long as the DG models in the sequence-component frame can be formulated. However, the convergence of SPF is heavily dependent on the choice of initial conditions. In this paper the Homotopic and Runge-Kutta order four (RK4) method is used to increase the robustness of SPF method due to initial conditions problem. An IEEE benchmark system is used to test the validity and robustness of the proposed algorithm.

Unpredictable increase in power demands will overload the supply subsystems and insufficiently powered systems will suffer from instabilities, in which voltages drop below acceptable levels. Additional power sources are needed to satisfy the demand. Small capacity distributed generators (DGs) serve for this purpose well. One advantage of DGs is that they can be installed close to loads, so as to minimise loses. Optimum placements and sizing of DGs are critical to increase system voltages and to reduce loses. This will finally increase the overall system efficiency. This work exploits Quantum Evolutionary Algorithm (QEA) for the placements and sizing. This optimisation targets the cheapest generation cost. Quantum Evolutionary Algorithm is an Evolutionary Algorithm running on quantum computing, which works based on qubits and states superposition of quantum mechanics. Evolutionary algorithm with qubit representation has a better characteristic of diversity than classical approaches, since it can represent superposition of states.

Increasing of electric power consumption and electricity price are making customers more sensitive in addressing the issues. Therefore, the accuracy of the recording device power consumption (kWh-meter) becomes an absolute necessity to reduce potential conflicts that may arise. This paper proposed prototype of smart-meter which combines transient peak value and steady state values to identify an activity of electrical appliances. These values are used as the identity of electrical appliances that will be taught to Constructive Backpropagation Neural Network (CBP-NN) to record power consumption in detail, including type appliance and time use. The proposed method has very simple structure, it only uses two input (transient peak value and steady state values) and single hidden layer with five neuron. The number of output is equal to the number of appliance. So that, the proposed method implement in microsprocessor system or in standalone product. Simulation and experimental results have validated the performance of the proposed method to operate in a real system

On line monitoring in distribution system suppose to keep the operation safe and reliable. It is connected measuring sensor that placed in nodes. To minimize great cost of sensors placement, state estimator is needed. This paper proposes online state estimator using neural network. Neural network distribution state estimation solves voltage estimation by using learning approach from power flow patterns. K-matrix distribution power flow is used as analysis method. Detailed information and position of network utility is displayed by Geographic Information System (GIS), control can easily do. NNDSE was design and test for single and three phase network. The results show that NNDSE reduce sensor almost 50%.

The proliferation of distributed generators (DGs) has altered a distribution network from a passive system to an active one. Therefore, power flow originally developed for the passive distribution system needs to account for a DG unit, which can operate either in PV or PQ control mode. Nevertheless, modifications previously proposed can only handle a limited number of PV buses. This paper presents a new robust three-phase distribution power flow, which incorporates the particle swarm optimization method into a Forward-Backward type method to handle a radial system with high R/X and high number of PV buses. Simulation and experimental results have validated the robustness and suitability of the proposed method to operate in a real microgrid system.

The interconnection of Distributed Generations (DG's) directly to the existing distribution systems has changed the characteristics of a distribution network from a passive network to an active one. This paper introduces a three phase load flow method that can handle both passive and active unbalanced radial distribution networks. The proposed method combines the direct-ZBR method (Forward-Backward based method) and Modified Lambda iteration. The direct ZBR load flow [1]is known as a fast and accurate method for solving passive radial three phase distribution systems. The Modified Lambda approach is chosen because of its simple algorithm that can be easily combined with direct ZBR method in handling DG'swhich are operated as voltage controlled(PV) buses. (6 pages)

The proliferation of distributed generators (DGs) and the concept of microgrids have altered a distribution network from a passive network to an active one. Hence, active distribution power flow methods need to account for a DG unit, which can operate either as a PV or PQ bus. However, some of the existing active distribution power flow methods have difficulty to converge if the resistance-to-reactance ratio (or R/X) is high. This paper presents a new robust three-phase distribution power flow, which incorporates the particle swarm optimization method into an existing distribution power flow to overcome these problems. An IEEE benchmark system is used to test the validity and robustness of the proposed algorithm.