Optimal tuning of PI speed controller coefficients for electric drives using neural network and genetic algorithms

This paper presents a method of tuning Proportional Integral (PI) controller coefficients in the off-line control of a nonlinear system. In this method, the first step is the identification of the system via Artificial Neural Networks (ANNs), using maximum overshoot and settling time obtained from the application circuit for different K_p-K_i pairs. With this in mind, multi-layer ANN, which uses back-propagation of the error algorithm, was used as the learning algorithm. In the second step, the purpose is to find the optimum controller coefficients using the ANN model as the objective function via Genetic Algorithms (GAs). A Digital Signal Processor (DSP-TMS320C50) was used to carry out control applications. The C++ language was used for ANN and GA, and and the Assembly language was used for the DSP. It is determined that maximum overshoot and settling time are very small if the system is controlled by control parameters obtained from the optimization process that uses GA.     

Optimum design of slurry pipelines by genetic algorithm

Various applications of slurry transportation through pipelines exist. A transportation problem is formulated to determine the pipe diameters and amounts of transported slurry from the production to the consumption points. The minimisation of the cost consisting of the pipe and energy cost terms is considered as the objective function. Pipe cost is given as the function of pipe diameters and the energy cost is defined as the function of pipe diameters and slurry amounts. Energy cost is obtained by using the relation that is previously determined after the experimental studies have been made for the magnetite ore. Genetic algorithm is used as the optimisation method and to apply this method a commercially available software written in the C language is used and modified. The proposed methodology to solve this non-linear programming problem is applied to a transportation system and it is found that the proposed methodology has made the complex, labour-intensive solution process very convenient for the users.     

Voltage-mode multifunction filter with mutually independent Q and \omega _0 control feature using VDDDAs

The paper focuses on the application possibilities of the newly presented voltage differencing active building block called voltage differencing differential difference amplifier. Using this active element, a multifunction frequency filter is designed featuring the possibility of mutually independent control of quality factor Q and characteristic frequency \(\omega _0\) by means of active elements. The structure of the filter is based on the idea of the Akerberg-Mossberg (AM) filter, i.e. the integrators in the structure are always realized only by two active elements. This fact results in better phase compensation for the filter. Compared to the AM opamp based filter, the newly proposed structure features high-impedance inputs, low-impedance output, and all basic frequency responses. The performance of the proposed structure has been verified by SPICE simulations using the TSMC \(0.18\,\upmu \hbox {m}\) level-7 SCN018 CMOS process parameters with \(\pm 0.9\,\hbox {V}\) supply voltage.     

Bio‐Hybrid Cell‐Based Actuators for Microsystems

As we move towards the miniaturization of devices to perform tasks at the nano and microscale, it has become increasingly important to develop new methods for actuation, sensing, and control. Over the past decade, bio‐hybrid methods have been investigated as a promising new approach to overcome the challenges of scaling down robotic and other functional devices. These methods integrate biological cells with artificial components and therefore, can take advantage of the intrinsic actuation and sensing functionalities of biological cells. Here, the recent advancements in bio‐hybrid actuation are reviewed, and the challenges associated with the design, fabrication, and control of bio‐hybrid microsystems are discussed. As a case study, focus is put on the development of bacteria‐driven microswimmers, which has been investigated as a targeted drug delivery carrier. Finally, a future outlook for the development of these systems is provided. The continued integration of biological and artificial components is envisioned to enable the performance of tasks at a smaller and smaller scale in the future, leading to the parallel and distributed operation of functional systems at the microscale.     

Prediction of surface roughness in abrasive waterjet machining of particle reinforced MMCs using genetic expression programming

Machining of particle-reinforced metal matrix composites has been considerably difficult due to the extremely abrasive nature of the reinforcements that causes rapid tool wear and high machining cost. Abrasive water jet (AWJ) machining has proven to be a viable technique to machine such materials compared to conventional machining processes. The present study is focused on the surface roughness of AWJ cut surfaces and genetic expression programming (GEP) was proposed to predict surface roughness in AWJ machining of 7075 Al alloy composites reinforced with Al₂O₃ particles. In the development predictive models, characteristics of materials such as size and weight fraction of reinforcement particles, and depth of cut were considered as model variables. The training and testing data sets were obtained from the well-established machining test results. The weight fraction of particle, size of particle, and depth of cut were used as independent input variables, while arithmetic mean of surface roughness, maximum roughness of profile height, and mean spacing of profile irregularity as dependent output variables. Different models for the output variables were predicted on the basis of training data set using GEP and accuracy of the best model was proved with testing data set. The test results showed that output variables increased with increasing input variables. The predicted results were compared with experimental results and found to be in good agreement with the experimentally observed ones.     

Computational investigation of testing parameter effects on abrasive wear behaviour of AL2O3 particle-reinforced MMCS using statistical analysis

The wear rate model of 7.3?vol.% Al₂O₃ particle-reinforced aluminium alloy composites with 16 and 66???m particle sizes fabricated by molten metal mixing method was developed in terms of applied load, particle size of reinforcement, abrasive grain size and sliding distance based on the Taguchi method. The two-body abrasive wear behaviour of the specimens was investigated using a pin-on-disc abrasion test apparatus where the sample slid against different SiC abrasives under the loads of 2 and 5?N at the room conditions. The orthogonal array, signal-to-noise ratio and analysis of variance were employed to find out the optimal testing parameters. The test results showed that particle size of reinforcement was found to be the most effective factor among the other control parameters on abrasive wear, followed by abrasive grain size. Moreover, the optimal combination of the testing parameters was determined and predicted. The predicted wear rate results were compared with experimental results and found to be quite reliable.     

Ge-Si-O phase separation and Ge nanocrystal growth in Ge:SiO(x)/SiO(2) multilayers--a new dc magnetron approach

Ge:SiO(x)/SiO(2) multilayers are fabricated using a new reactive dc magnetron sputtering approach. The influence of the multilayer stoichiometry on the ternary Ge-Si-O phase separation and the subsequent size-controlled Ge nanocrystal formation is explored by means of x-ray absorption spectroscopy, x-ray diffraction, electron microscopy and Raman spectroscopy. The ternary system Ge-Si-O reveals complete Ge-O phase separation at 400?°C which does not differ significantly to the binary Ge-O system. Ge nanocrystals of < 5?nm size are generated after subsequent annealing below 700?°C. It is shown that Ge oxides contained in the as-deposited multilayers are reduced by a surrounding unsaturated silica matrix. A stoichiometric regime was found where almost no GeO(2) is present after annealing. Thus, the Ge nanocrystals become completely embedded in a stoichiometric silica matrix favouring the use for photovoltaic applications.     

Some structural properties of CdO:F films produced by ultrasonic spray pyrolysis method

The fluorine doped cadmium oxide (CdO:F) samples have been deposited at 250 °C by ultrasonic spray pyrolysis method. Cadmiumacetat-dihydrat and ammonium fluoride have been taken as a source of cadmium and fluorine-dopant respectively. The thickness of the CdO:F samples was about 1.4 μm. X-ray diffraction pattern of the CdO:F samples has revealed that the samples are polycrystalline with cubic sodium chloride structure. There are shifts of the d values (interplanar spacing) for CdO:F samples with respect to standard CdO film. The lattice parameters for cubic structure have been calculated using the Bragg equation. The texture coefficients calculated for various planes at different fluorine concentrations indicate that the samples have exhibited (111) and (200) preferential orientations.     

Effects of fluorine doping on the structural properties of the CdO films deposited by ultrasonic spray pyrolysis

The fluorine doped cadmium oxide samples have been deposited at 250 °C by ultrasonic spray pyrolysis method. X-ray diffraction patterns of the CdO:F samples have revealed that the samples are polycrystalline with cubic sodium chloride structure. The texture coefficients calculated for various planes at different fluorine concentrations indicate that the samples have exhibited (1 1 1) and (2 0 0) preferential orientations. The lattice parameters for cubic structure of each diffraction plane have been calculated. The crystallite size of the samples being nearly constant until 4% of fluorine doping showed reasonable decrease above this concentration value. The macro strain and dislocation density vary with fluorine concentrations.