Hasil Pencarian  ::  Simpan CSV :: Kembali

"Neural networks and fuzzy systems are model free control design approaches that represent an advantage over classical control when dealing with complicated nonlinear actuator dynamics. Neuro-Fuzzy Control of Industrial Systems with Actuator Nonlinearities brings neural networks and fuzzy logic together with dynamical control systems. Each chapter presents powerful control approaches for the design of intelligent controllers to compensate for actuator nonlinearities such as time delay, friction, deadzone, and backlash that can be found in all industrial motion systems, plus a thorough development, rigorous stability proofs, and simulation examples for each design. In the final chapter, the authors develop a framework to implement intelligent control schemes on actual systems. Rigorous stability proofs are further verified by computer simulations, and appendices contain the computer code needed to build intelligent controllers for real-time applications. Neural networks capture the parallel processing and learning capabilities of biological nervous systems, and fuzzy logic captures the decision-making capabilities of human linguistics and cognitive systems."

"ABSTRAKIndustri kemasan adalah sektor industri yang diproyeksi akan terus bertumbuh, khususnya industri kemasan plastik. Dengan plastik merupakan turunan dari minyak mentah, plastik dan bahan bakunya yang berupa resin tergolong sebagai produk petrokimia. Harga resin berfluktuasi dan sensitif terhadap pergerakan harga minyak mentah sehingga deret waktunya bersifat nonstasioner. Kondisi ini dapat membuat konverter plastik sebagai pihak yang mengkonversi resin menjadi produk plastik mengalami kesulitan dalam memasingkan harga resin pada harga produk kepada pelanggannya serta berpotensi mengalami kerugian. Hal ini memicu peneliti untuk memodelkan peramalan harga resin. Literatur menunjukkan peramalan harga produk petrokimia dengan model tradisional maupun komputasi lunak, namun masih memiliki keterbatasan yang berefek pada akurasi peramalan. Selain itu, kebanyakan peneliti memodelkan peramalan pada harga minyak mentah dan tidak ada yang ditemukan menggunakan harga resin sebagai objek peramalan. Penelitian ini mengajukan peramalan harga resin dengan model neuro-fuzzy yaitu ANFIS dan membandingkan hasilnya dengan sebuah model tradisional, ARIMA, dan sebuah model tunggal komputasi lunak, NN. Hasil peramalan menunjukkan model ANFIS memiliki tingkat error dalam bentuk MAPE yang relatif sangat kecil yaitu 1.06% dan juga tingkat akurasi arah yang tinggi yaitu 93%. Ini menunjukkan bahwa model peramalan ANFIS dapat merepresentasikan karakteristik harga resin. Selain itu, hasil uji statistik juga menunjukkan bahwa terdapat perbedaan yang signifikan antara akurasi peramalan ANFIS jika dibandingkan dengan ARIMA dan NN.

---

ABSTRACTThe packaging industry is an industrial sector that is projected to continue growing, especially the plastic packaging industry.While plastic is a derivative of crude oil, plastic and its raw material which is called resin are categorized as petrochemical products. The resin price is fluctuating and sensitive to crude oil price movement so that the time series is nonstationary. This condition may cause the plastic converters as the ones converting resin to plastic products to experience the difficulty in passing the resin price into the product price for the customers and have the potential to suffer loss. This triggers the researcher to model the forecasting of resin price. Literatures show petrochemical product price forecasting by using traditional as well as soft computing models, but they still have limitations that affect the forecasting accuracy. In addition, most researchers model forecasting on crude oil price and none of them found to use the resin price as forecasting object. This research proposes resin price forecasting using neuro-fuzzy model that is ANFIS and compare the result with a traditional model, ARIMA, and a standalone soft computing model, NN. Forecasting result shows that ANFIS model has a relative low error in terms of MAPE that is 1.06% and also a high directional accuracy which is 93%. This shows that ANFIS forecasting model can represent the resin price characteristic. Moreover, statistical test also shows that there is significant difference in ANFIS forecasting accuracy if compared to ARIMA and NN."

"Neuro fuzzy system has been shown to provide a good performance on chromosome classification but does not ofter a simple methods to obtain the accurate parameter values required to yield the best recognition rate....."

"Identifikasi retina merupakan metode identifikasi biometrik dengan tingkat kesalahan rendah melalui pola-pola unik pembuluh darah di bagian belakang retina. Pola-pola ini dapat digunakan sebagai data latih logika neuro fuzzy untuk kemudian digunakan sebagai pembanding pada saat identifikasi dilakukan.Penelitian ini bertujuan untuk mengenali citra retina mata manusia, baik bagian kiri maupun kanan, menggunakan teknik pengolahan citra dan Adaptive Neuro Fuzzy Inference System (ANFIS). Pada proses pengenalan retina ini, citra digital yang sudah diakuisisi akan dicrop dan dibagi menjadi image block berukuran 4x4. Kemudian blok citra dikonversi dari format Red Green Blue (RGB) menjadi format Hue Saturation Value (HSV). Untuk mendapatkan parameter fitur warna HSV, setiap komponen warna HSV dihitung nilai rata-ratanya. Nilai rata-rata HSV dimasukkan ke dalam database dan dilatih dengan ANFIS yang terdiri atas 2 jenis membership function, yaitu Gaussian dan Trapesium dengan 3 input dan 1 ouput.Dari hasil uji coba, hasil identifikasi memiliki tingkat akurasi hingga 65% untuk membership function Trapesium dan 80% untuk membership function Gaussian dengan 60 kali pelatihan ANFIS.

---

Retina identification is a biometric identification method which has very low error rate using a unique blood vessel pattern in the back of the retina. The identification involved an infrared scanned retina imagery which is analyzed using image processing technique to derive the color characteristics and then trained into the Adaptive Neuro Fuzzy Inference System (ANFIS).The objective of this research to identify a person?s identity from his/her retina image. The identification process is started by cropping the digital retina image then transformed into an 4x4 image block. The image block is then converted from Red Green Blue (RGB) color format to the Hue Saturation Value (HSV) format. Each color component of HSV values is then averaged, saved to a database and trained using ANFIS. The Neuro fuzzy used Gaussian and Trapezoid membership function which have 3 input and 1 ouput, respectively.The simulation results showed the identification system has an accuracy rate up to 65% and up to 80%, for Trapezoid and Gaussian membership function, respectively. This results are achieved using 60 training data in the ANFIS."

"This book covers algorithmic and hardware implementation techniques to enable embedded deep learning. The authors describe synergetic design approaches on the application, algorithmic, computer architecture, and circuit-level that will help in achieving the goal of reducing the computational cost of deep learning algorithms. The impact of these techniques is displayed in four silicon prototypes for embedded deep learning.Gives a wide overview of a series of effective solutions for energy efficient neural networks on battery constrained wearable devices;Discusses the optimization of neural networks for embedded deployment on all levels of the design hierarchy-applications, algorithms, hardware architectures, and circuits-supported by real silicon prototypes;Elaborates on how to design efficient Convolutional Neural Network processors, exploiting parallelism and data-reuse, sparse operations, and low-precision computations;Supports the introduced theory and design concepts by four real silicon prototypes. The physical realizations implementation and achieved performances are discussed elaborately to illustrated and highlight the introduced cross-layer design concepts."

"Metode Adaptif Neuro Fuzzy Inference System (ANFIS) untuk penerapan pada identifikasi dan aplikasi kendali sistem multi masukan multi keluaran (MIMO) dan sistem satu masukan satu keluaran (SISO) diharapkan dapat menjadi salah satu metode kendali cerdas alternatif selain mengandalkan metode kendali cerdas umum seperti Jaringan Syaraf Tiruan backpropagation. Sistem plant MIMO tersebut mengacu kepada sistem Pesawat Udara Nirawak SRITI yang menghasilkan 3 surface kendali. Metode ANFIS yang dibangun merupakan metode yang terdiri dari metode Jaringan Syaraf Tiruan Adaptif dan model sistem inferensi fuzzy. Algoritma pembelajaran identifikasi, invers, dan algoritma pembelajaran On-Line merupakan metode pembelajaran yang digunakan pada sistem ini. Melalui rancangan metode ANFIS ini kemudian dilakukan simulasi untuk memperlihatkan hasil identifikasi dan pembelajaran secara On-line sistem ketika masukan dan keluaran sistem Pesawat Udara Nirawak (UAV) diberikan. ANFIS dengan algoritma pembelajaran identifikasi dan invers telah dapat memberikan hasil respon yang baik, namun untuk menyempurnakan hasil metode pembelajaran Off-line sistem harus diberikan suatu pengestimasi tambahan yang menjadikannya sistem On-line. Hasil percobaan On-line telah menunjukkan keberhasilan sistem ANFIS dalam mengidentifikasi dan mempelajari sistem SISO dan MIMO.

---

Adaptif Neuro Fuzzy Inference System (ANFIS) method for Multi Input Multi Output (MIMO) plant system identification and control application expected to become one of an alternative smart control method in addition to relying on another smart control method such as backpropagation neural network. That MIMO plant system refers to Unmanned AeroVehicle which produce 3 control surface.

ANFIS method which will be proposed consist of adaptive neural network method and Fuzzy Inference System model. Identification learning algorithm, inverse learning algorithm, and On-line learning are identification and control methods used in this system.

From this proposed ANFIS method then simulated to demonstrate the identification and learning’s output when UAV SRITI plant system's input and output were given. ANFIS with identification and inverse learning algorithm had given good response, but for more perfection of Off-line system method, there should be given some additional estimator to make it On-line. The On-line method result has demonstrated the success of ANFIS system in identifying and learning SISO and MIMO systems."

"Pada masa sekarang ini perkembangan teknologi cenderung memiliki kemampuan untuk berpikir dan mengambil keputusan layaknya manusia. Salah satu dari banyak metode untuk mengembangkan teknologi yang cerdas adalah dengan menggunakan Adaptive Neuro Fuzyy Inference System. Penelitian ini dilakukan dengan menerapkan ANFIS tipe Sugeno pada data-data penelitian umum, seperti data tanaman iris dan data ionosphere, melihat efek perubahan parameter-parameter terhadap recognisinya, lalu melakukan ANFIS terhadap data citra wajah.

---

The technology nowadays tends to have abbility to think and to size up decision, just like us humans. One of the kind of method to enhance smart technology is by using Adaptive Neuro Fuzyy Inference System. This research is done by using ANFIS Sugeno type on general research data, such as iris plant data and ionosphere data, observing the effect of the changing parameter over the recognition, then using ANFIS on face image data."