The use of multi-agent architecture allows the development of a control system for autonomous navigation of mobile robots. So, in this study, a new method based on the paradigm of a multi-agent system have been proposed in order to control the robot. The objective of this work is to create collision-free trajectories for mobile robot to reach the goal position. The main idea of this approach is based on the potential fields method, the robot is attracted to the target and pushed from the obstacles. Therefore, the proposed controller architecture is composed of three agents: the robot navigation agent helps to control the robot to the target location, and it is implemented with a fuzzy based controller. The obstacle avoidance agent is responsible for keeping the robot away from obstacles, this agent is developed using a hierarchical fuzzy controller. This is to reduce the number of fuzzy rules according to the number of fuzzy input variables. Only one perception agent for eight sensors in the robot provides the necessary information about the environment. The motivation of using less number of agents is to reduce conflict situations between agents, and therefore coordination process will be needed with minimum communication overhead. For the coordination method, an utility function in order to remove conflict situations among agents has proposed. Finally, the simulation results confirms the effectiveness and efficiency of the proposed approach is presented, so that the robot can reach the target position without encountering obstacles in an environment with various obstacles. It also shows that the robot is not able to achieve the target's position alone by the agent of the robot navigation and the avoidance barrier. But with the proposed synchronization method of these two factors, it is possible to achieve target’s position without dealing with obstacles for the robot.

Unbalance and misalignment are major concerns in rotating machinery.The vibrations caused by misalignment and Unbalance may destroy critical parts of the machine, such as bearings, seals, gears, and couplings. Statistics show that the vibration level have risen in the rotating machinery is the main cause of unbalance, about 80 percent.In this study, the diagnosis , severity,type and localization of the unbalance using intelligent methods discussed also this research describes the aplication of this method for the prediction of the effect of combined faults of unbalance and other fault such as misalignment and belt fault on the frequency components of vibration signal of the rotating machinery.Presented methods are consists of five stages, signal processing, feature calculation, feature extarction, feature selection and classifieing faults using support vector machine (SVM). the data are taken experimentally and translated to signals in frequency domains by the Fast Fourier Transform (FFT), are powerful in diagnosing a variety of vibration-related problems in rotating machinery. also by the integration of FFT and Envelope Analysis reduces the noise influence in signals. The hybrid method of FFT and Envelope Analysis has efficiently extracted the fault features.In this study, 16 statistical features are calculated from frequency domain as origin feauture.The ICA is used for feature extraction and data reduction from origin features. The principal components analysis(PCA) is also applied in feature extraction process for comparison with ICA does. After feature extraction, we performed feature selection process to remove irrelevant and useless feature. The distance evaluation technique was employed due to its simple and reliability. SVMs-based multi-class classification is applied to do faults classification process. The results show that SVMs achieved high performance in classification using one-against-all as multiclass strategy. in this study, we utilized a cross-validation technique using 10-fold in order to choose the optimal parameters of support vector machine and a Gaussian kernel function.using ICA feature extraction and combining kernel parameters selection gave the best faults classification. According to this result, the combination of ICA and SVMs can serve as a promising alternative for intelligent faults diagnosis in the future.

Many oil wells can send fluids to the ground by the natural pressure existing in reservoirs in the first stages of production. However, after some time, the production of wells are stopped or reduced due to the drop of pressure in reservoirs or other problems leading to the fact that friction forces become greater than internal pressure of wells. The newest model of structure of these pumps is Rotaflex which have a higher efficiency compared with other models. The generated energy in surface equipment is transferred to subsurface equipment by a rod string. Due to the elasticity of this rod string, its stroke of the surface and subsurface sections are different. In this study, effort was made to model the elastic behavior of rod string and consider the effects of fluid inertia to calculate parameters such as subsurface stroke, change in the length of tubing, force on the polished rod and plunger and fluid pressure changes. To this end, rod string damped wave equation, continuity equations, fluid momentum and the equation of changes in pump chamber pressure were extracted first. Then, these equations were discretized through Finite Difference Method; and finally, non-linear coupled equations were solved regarding the data of a specific well using Newton-Raphson Method. The results are calculating surface and subsurface stroke, the effect of fluid pumping speed, plunger diameter change on change in the length of tubing, calculating counterbalance effect and calculating the pump pressure changes and the discharge pressure changes

Misalignment in rotating machines is one of the most important defects seen in the industry.The disadvantage of the high percentage of failure of rotating equipment such as couplings and bearing to be allocated. In this study, the diagnosis and severity of the misalignment using intelligent methods discussed. Troubleshooting methods are provided, that include the steps of signal processing, feature extraction, feature selection and classification of defects. In this study, simultaneous feature selection and classification of defects and with a combination of particle swarm optimization and support vector machine algorithm is performed. Signal processing using both continuous and discrete wavelet transform is performed. mother wavelet, bior3.1, based on maximum energy on Shannon entropy ratio, as best wavelet, from among 211 mother wavelet was chosen. In the continuous wavelet transform with maximum energy to entropy ratio, scales from 80 to 128 were considered as optimal scales. Similarly, the discrete wavelet transform is also used this method to analyze the fifth level signal is selected. In this study of 40 features (18 statistical features and 22 features a maximum Likelihood Estimation) was used. Kurtosis and mean harmonic features in continuous wavelet analysis method, and mean, moment and maximum Likelihood Estimation with functions of parito, lognormal and gamma features in discrete wavelet method, as best features in diagnosis and classification of defects were selected. To detect flaws In this study, the four support vector machine with one against all method, in each of continuous and discrete wavelet techniques have been used. In determining the severity of the misalignment, used one against one method for any mode misalignment. It should be noted that in this study, feature selection and determine the optimal parameters of support vector machine and a Gaussian kernel function, along with particle swarm algorithm, optimized.

 Gears are used widely in the power transmission system. cracks and other damages may occur in gears. Since large load bearing and power transmission, it is required to analyze the dynamic behavior of gears. By improving the reliability and maintenance of mechanical systems, a lot of people check gear damage detection. The modal analysis and investigation of dynamic characteristics is one way to check cracked gear. Ansys is used As a powerful software to analyze the vibrating systems. In this study, three-dimensional gear model has been checked and the natural frequencies of cracked gear system has been obtained and the effect of size and location of cracks on natural frequencies are investigated. It also uses neural network to find crack characteristic

 The propagation of ultrasonic Lamb and Shear Horizontal waves in a homogeneous and isotropic micro plate under the influence of microstructure size effects has been studied in this thesis. In the current study, it is assumed that the plate is infinite in its plane and the upper and lower surfaces of plate are free of any tension. To consider the impact of micro-scale effects on the propagation of waves, three different couple stress theories known as Indeterminate, Modified and Consistent couple stress theory, has been used to extract Dispersion equations. In these theories, a parameter called the characteristic length scale parameter is included to account for the size of the microstructure of materials. The aim of this study is to extract the phase and group velocity dispersion curves with respect to frequency and evaluate the impact of variations of characteristic length scale parameter on the behavior of these curves. In order to verify the solution method, first dispersion curves are plotted using classical theory of elasticity and the obtained results are validated by comparing with different references. The dimensionless dispersion equations are obtained using three types of couple stress theory and then solved by numerical methods using MATLAB software. Then the dispersion curves of phase and group velocities are plotted with respect to frequency. Since it has been observed that the propagation speed of waves significantly increases, it has been concluded that due to presence of character length scale parameter, the couple stress theories predict much stiffer material compared to classical theory. It has been observed that in Lamb wave dispersion curves, the value of wave velocity converges to speed of dispersive Rayleigh waves at high frequencies. In addition, Horizontal shear waves were observed to be ascentional. Also by experimental measurement of phase or group Lamb waves velocity with different frequencies and comparing with plotted dispersion curves, a novel method is proposed to find characteristic length parameter in the couple stress theories. This method not only can reduce the cost of testing, it also helps us to determine the precise properties of micro plates using ultrasonic waves propagation. The extracted equations in this study, not only can be used for micro plates, but also with plates in different dimentions, so that their bahaviour is influenced by their internal structure. From this materials, bones, some types of particle reinforced composites and polymers can be mentioned.

 Exact identification of structural behavior is necessary to provide High Reliability, less vibration and increasing fatigue life of the structures. In this study, power train shafts are investigated. Multisegment power train shafts with flexible couplings are affected by high vibration. First, the system will be modeled by finite element method and then the best points for hammer test are determined. Natural frequencies, mode shapes and damping coefficients of transverse vibration acquired by modal analysis. Inherently, there are modeling error in finite element method for modeling a system. These errors can be eliminated or reduced by model updating methods. Effectiveness of the model parameters on the natural frequencies are calculated by sensitivity analysis. To reduce finite element modeling of the shafts, an objective function is defined and then three optimization methods (Nelder-Mead, BA and PSO) are applied on shaft and flexible coupling model.

 Manufacturers are always looking for ways to reduce the cost of repairs and maintenance of their equipments. However, using new methods of maintenance and condition monitoring of machines, can prevent the occurrence of defects in the system and to overcome defects, repairs and maintenance costs reduced. In the air system due to human and safety factors, condition monitoring of critical component, especially rotating component, is very important. Helicopters are classified as rotary wing aircrafts. Therefore, the need to condition monitoring of critical component of helicopter, as health and usage monitoring system of helicopter is propound. Due to importance of the helicopter transmission, monitoring components of this part, is very important. The purpose of this project, is propose an algorithm to determine the helicopter tail shafts of shahed 278 misalignment. Therefore, the vibration signals measured by sensors mounted on the helicopter tail section bearings are collected. Then, the signal processing of continuous wavelet transform and discrete wavelet transform are used, and two sets of features, statistical features and maximum likelihood parameters, are extracted. Using the distance factor, the features that are more sensitive, are selected and as inputs to the neural network are considered. For classification, Generalized Regression Neural Network and radial basis function neural network are used. The results showed that the statistical feature of entropy of energy logarithmic feature and standard deviation of the distribution of the maximum likelihood, have a higher efficiency than other features. Also, Generalized Regression Neural Network has higher reliability capability

Since nowadays the transmit pipelines have been greatly developed and
kilometers of pipelines exist all over the world, maintenance are very important.
One of the major damages in the pipe line is leakage of the inside fluid. This would
cause the economic loss, environmental or even life damages. One way to detect
the leakage is acoustic emission testing. In this thesis, the leakage in the pipeline is
simulated using acoustic emission testing in the ABAQUS software. Simulating the
latter in the software has great importance, because it’s possible to investigate the
different conditions by altering the parameters in the software instead of impossible
or expensive tests. Because there aren’t adequate information about the basis of the
leakage, it was modeled using three different functions. Signals of the different
leakage functions are recorded on the two sides of the investigating region through
the pipe using two sensors; and are estimated by the correlation function. It was
observed that as the leakage location is getting closer to the center of the pipe on
investigating region, the error is reduced. By imposing the window on the
receiving signals, the results are improved and error is decreased. Finally, in order
to demonstrate validity of the results, modeling is compared to the practical ones
and recommendations for the further investigations are presented.

 In the treatment of leukemia, after Chemotherapy drugs, the patient's immune system weakened. So, a certain period of time is determined for return the patient to normal, The chemotherapy protocol is a set of a preset, where the exact amount of time and deliver chemotherapy after chemotherapy is given, Drugs as safe drug therapy or supportive therapy to accelerate the healing process is injected into the patient's body. Granulocyte stimulating factor is one of these drugs which cause strengthening the immune system, in this research, particular chemotherapy protocols are optimized by Reduction of immunotherapy And Reducing injection of immunotherapy drug. To studying human body, it is necessary to use a mathematical model. This model, has to modeling the effect of drugs on Different stages of cell growth. By proper modeling of the effects of drugs, without direct testing, will take a big step in discovering low-cost method of treatment. Scholz model was used in this research. The model was simulating in Matlab software. In this way, the objective function which is the integral of the difference between the response of the model and the experimental results will fall 26/81%. Changes in intermediate stages of cell growth, all within the scope of the changes were real. To help optimize the timing and rate of growth granulocyte colony stimulating factor, transfusions have been investigated, also change the program of Injection and check the response of the model during resuscitation, the CHOEP protocol from 14 days to 11 days decreased. In addition to reducing immunotherapy period from 10 days to 6 days, the total dose of 10 units during the period dropped to 2.35 units.

The corrosion is the main reason causing leaks in the pipes. To prevent the continuation of the economic costs of meeting various environmental hazards, detection leak and fix it, is necessary and crucial. There are two main ways for ensuring of piplines performance: destructive and non-destructive inspection. Acoustic emission is a non-destructive leakage testing of pipeline. The advantage of this method is that the tanks and pressurized system can without shutting down the during the service shall be periodically reassessed and revised. Jet exhaust by leak causes acoustic waves and vibration lead to its radius displacement. In this project, modeling of AE waves due to vibration resulting from the leakage are discussed. Donnell Theory has been used for analyzing of non-line dynamic of pipe and potential function for the interaction of fluid flow and pipe. Galerkin method used for the discretization of pipe radial displacement equation and nonlinear differential equations whit 7 degrees of freedom is derived. To solve differential equations used MATLAB software. The results of theory with practical results were compared and wavelet packet transform of acoustic emission waves from leaking is taken. The results indicate that acoustic waves of leak have frequency range 200-300 kHz.

هر صنعتی نیازمند برنامه های پیشگویانه، در جهت بهینه نمودن مدیریت منابع و بهبود اقتصاد کارخانه با کاهش هزینه های غیر ضروری و افزایش سطح ایمنی می باشد. درصد بسیاری از خرابی های موجود در کارخانجات در فرآیند تولید، به دلیل خرابی یاتاقان های غلتشی میباشد. هدف از این پژوهش پیشنهاد الگوریتمی، برپایهی روش تجزیه ی مقادیر منفرد، و به کارگیری روشهایی در زمینه عیبیابی یاتاقان های غلتشی می باشد. روش های عیب یابی ارائه شده دارای چهار مرحله ی پردازش سیگنال، استخراج ویژگی، انتخاب ویژگی و دسته بندی عیوب با استفاده از شبکه های عصبی مصنوعی می باشند. از شش روش پردازش سیگنالی الگوریتم تجزیه ی مقادیر منفرد، تحلیل پوش، تبدیل موجک پیوسته، تبدیل موجک گسسته و ترکیب تبدیل موجک پیوسته با روش های تحلیل پوش و تجزیه ی مقادیر منفرد، جهت سه مجموعه داده ی یاتاقان های غلتشی، استفاده شده است. موجک های مناسب با استفاده از سه معیار بیشترین انرژی، کمترین آنتروپی و بیشترین نسبت انرژی به آنتروپی از میان 87 موجک مادر بررسی شده، انتخاب شده اند. موجک بایر 1.3 با استفاده از معیار انتخاب موجک انرژی، به عنوان بهترین موجک در زمینه ی عیب یابی انتخاب گردید. در این پژوهش از دوازده ویژگی آماری و چهار روش انتخاب ویژگی استفاده شده است. ویژگی های میانه، انحراف معیار استاندارد و ویژگی ترکیبی دوازدهم، دارای کارائی بهتری نسبت به ویژگی های بررسی شده در پژوهش، در تشخیص عیب یاتاقان غلتشی می باشند. همچنین معیار تفکیک کننده خطی فیشر، معیار بهتری برای انتخاب ویژگی ها نسبت به سایر ویژگی های بررسی شده در این پژوهش می باشد. شبکه ی شعاع مبنا نیز دارای قابلیت اطمینان بالاتری نسبت به سایر شبکه های بررسی شده در این کاربرد است. همچنین الگوریتمی برمبنای تجزیه ی مقادیر منفرد ارائه گردیده است که نتایج، نشان دهنده ی کارائی بالای این روش، برای تشخیص عیوب یاتاقان های سرعت بالا و سرعت پائین می باشد.