Morteza Mohammadzaheri - Academia.edu (original) (raw)

Papers by Morteza Mohammadzaheri

AIP advances, Jun 1, 2024

Research Square (Research Square), Apr 5, 2024

Accurate and timely diagnosis of dementia progression remains a major global challenge due to the... more Accurate and timely diagnosis of dementia progression remains a major global challenge due to the complexities of brain pathology and the lack of definitive biomarkers. This study presents a pioneering fully connected cascade (FCC) neural network model that leverages cost-effective lifestyle and neuroimaging data to predict dementia progression with remarkable accuracy. The model uniquely integrates 42 lifestyle factors for brain health (LIBRA) and 7 brain atrophy and lesion indice (BALI) derived from baseline MRI data as inputs, to predict sensitive diffusion tensor imaging (DTI) biomarkers of white matter degeneration. Remarkably, the FCC network achieved a mean squared error of 0.0071693 in predicting DTI metrics, demonstrating exceptional predictive capability. This multidisciplinary data-driven approach capitalizes on the model's ability to detect subtle yet informative changes in brain structure and function through advanced neuroimaging. By amalgamating multidomain lifestyle and neuroimaging data, the proposed model enhances diagnostic value and sensitivity to dementia pathology. Its high accuracy, scalability with large datasets, clinical interpretability, and cost-effectiveness make it a powerful computational tool for early prediction, monitoring, and personalized treatment planning in dementia care. This groundbreaking research exemplifies the transformative potential of artificial intelligence in tackling the global dementia burden, paving the way for improved patient outcomes and reduced healthcare costs.

Research Square (Research Square), Apr 2, 2024

Early prediction of dementia and disease progression remains challenging. This study presents a n... more Early prediction of dementia and disease progression remains challenging. This study presents a novel machine learning framework for dementia diagnosis by integrating multimodal neuroimaging biomarkers and inexpensive, readily available clinical factors. Fractional anisotropy (FA) measurements in diffusion tensor imaging (DTI) provide microstructural insights into white matter integrity disturbances in dementia. However, acquiring DTI is costly and time-consuming. We applied Recursive Feature Elimination (RFE) to identify predictors from structural measures of the 9 Brain Atrophy and Lesion Index (BALI) factors and 42 Clinical Lifestyle for Brain Health (LIBRA) factors to estimate fractional anisotropy (FA) in DTI. The 10 most effective BALI/LIBRA features selected by RFE were used to train an interpretable decision tree model to predict dementia severity from DTI. A decision tree model based on biomarkers selected by Recursive Feature Elimination (RFE) achieved an accuracy of 96.25% in predicting dementia in an independent test set. This integrated framework pioneers the prediction of white matter microstructural changes from available structural/clinical factors using machine learning. By avoiding DTI acquisition, our approach provides a practical and objective tool to enhance dementia screening and progress monitoring. Identification of key predictive markers of BALI/LIBRA will also provide insights into lifestyle-related disease mechanisms, neurodegeneration, and white matter dysfunction.

Asian Journal of Control, Aug 2, 2011

In this review article, the most popular types of neural network control systems are briefly intr... more In this review article, the most popular types of neural network control systems are briefly introduced and their main features are reviewed. Neuro control systems are defined as control systems in which at least one artificial neural network (ANN) is directly involved in generating the control command. Initially, neural networks were mostly used to model system dynamics inversely to produce a control command which pushes the system towards a desired or reference value of the output (1989). At the next stage, neural networks were trained to track a reference model, and ANN model reference control appeared (1990). In that method, ANNs were used to extend the application of adaptive reference model control, which was a well‐known control technique. This attitude towards the extension of the application of well‐known control methods using ANNs was followed by the development of ANN model‐predictive (1991), ANN sliding mode (1994) and ANN feedback linearization (1995) techniques. As the first category of neuro controllers, inverse dynamics ANN controllers were frequently used to form a control system together with other controllers, but this attitude faded as other types of ANN control systems were developed. However, recently, this approach has been revived. In the last decade, control system designers started to use ANNs to compensate/cancel undesired or uncertain parts of systems' dynamics to facilitate the use of well‐known conventional control systems. The resultant control system usually includes two or three controllers. In this paper, applications of different ANN control systems are also addressed.Copyright © 2011 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society

Asian Journal of Control, Feb 28, 2011

In this paper, intelligent control of a particular Catalytic Continuous Stirred Tank Reactor (CST... more In this paper, intelligent control of a particular Catalytic Continuous Stirred Tank Reactor (CSTR) is addressed. Control command is the sum of two components: steady state and transient commands. A fuzzy controller generates a transient control command that pushes the system towards the setpoint. A steady state control command is generated to maintain the steady state situation at the setpoint (based on the concept of 'control equilibrium point'). For comparison, a case study was also controlled by the neuro-predictive algorithm. The studied CSTR is known as a good example of a neuro-predictive control application; however, the newly proposed hybrid intelligent method leads to much better setpoint tracking as well as less change in control command (which is very important in implementation of the control system).

The control methods used for shock or free vibration suppression are usually different from those... more The control methods used for shock or free vibration suppression are usually different from those for forced vibration cancellation, because shock vibration can be regarded as a type of transient vibration that is different in nature from steady-state forced vibration. In reality, however, both steady-state and transient excitations may occur in flexible structures, so there is a need to control both types of vibration. To show the integration of various vibration control strategies, a hybrid control system based on adhesive strain gauges and PZT patches is proposed to construct a distributed resonant absorber and a distributed shock absorber together. The hybrid system is governed by a control arbitrator that switches between the two active vibration absorbers according to the different excitation conditions. The effectiveness of the integrated system is shown through simulation and experimental studies.Ley Chen, Morteza Mohammadzaherihttp://www.aus.edu/conferences/isma07/docs/Booklet%20layout-2007_20070326.pd

ABSTRACT A number of models have been presented to estimate the displacement of piezoelectric act... more ABSTRACT A number of models have been presented to estimate the displacement of piezoelectric actuators; these models remove the need for accurate displacement sensors used in nanopositioning. Physics based models, inspired by physical phenomena, are widely used for this purpose due to their accuracy and comparatively low number of parameters. The common issue of these models is the lack of a non-ad-hoc and reliable method to estimate their parameters. Parameter identification of a widely accepted physics-based model, introduced by Voigt, is addressed in this paper. Non-linear governing equation of this model consists of five parameters needing to be identified. This research aims at developing/adopting an optimal and standard (non-ad-hoc) parameter identification algorithm to accurately determine the parameters of the model and, in a more general view, all physics-based models of piezoelectric actuators. In this paper, Genetic Algorithm (GA) which is a global optimisation method is employed to identify the model parameters.

ABSTRACT This article addresses sensorless control of a piezoelectric tube actuator to avoid the ... more ABSTRACT This article addresses sensorless control of a piezoelectric tube actuator to avoid the expense and practical limits of displacement sensors in nanopositioning. Three electrical signals have been used to estimate displacement: the piezoelectric voltage, the induced voltage and the sensing voltage. In this work, the piezoelectric voltage was employed to estimate displacement which does not require drift removal like the sensing voltage and does not suffer from a time lag respect to displacement like the induced voltage. This signal is the actuating signal at the same time, so the sensorless control system is feedforward. It was shown the relationship between the piezoelectric actuator and displacement is nearly linear at the designated operation area: excitation of the tube by triangular voltage functions with the magnitude up to 60V and the frequency up to 60Hz. Therefore, Internal Model Control (IMC) was employed to design this feedforward controller based on a second order linear discrete model which maps the piezoelectric voltage into displacement. The performance of the proposed feedforward controller has been compared with a well-tuned feedforward P-action controller and a remarkable improvement has been observed.

International Communications in Heat and Mass Transfer, 2012

In this work, a new solution approach was developed for heat estimation class of inverse heat tra... more In this work, a new solution approach was developed for heat estimation class of inverse heat transfer problems where radiation provides the dominant mode thermal energy transport. An Artificial Neural Network (ANN) was designed, trained and employed to estimate the heat emitted to irradiative batch drying process. In a simple laboratory drying furnace, various input signals (different input power functions) were input to the dryer's halogen lamp and the resulting temperature history were measured and recorded for a point on the bottom surface of the dryer. After estimating the order, the sampling time and the dead-time of the system, the recorded data were arranged for inverse modelling purposes. Next, an artificial neural network (ANN) was designed and trained to play the role of the inverse heat transfer model. The results showed that ANNs are applicable to solve inverse heat estimation problems of irradiative batch drying process. An important advantage of this method in comparison with classical inverse heat transfer modelling approaches, detailed knowledge of the geometrical and thermal properties of the system (such as wall conductivity, emissivity , etc.) is not necessary. Such properties are difficult to measure and may undergo significant changes during the temperature transient.

Asian Journal of Control, Jun 22, 2010

In this paper the concept of Control Inertia is introduced and based on this concept, unexpectedl... more In this paper the concept of Control Inertia is introduced and based on this concept, unexpectedly inadequate control behaviour of High Control Inertia systems is explained. Fuzzy compensators are then suggested to improve the control behaviour. This work is in the area of non-model-based control. In order to indicate the merit of the proposed technique, a neuro-predictive (NP) control is designed and implemented on a highly non-linear system, a lab helicopter, in a constrained situation. It is observed that the behaviour of the closed loop system under the NP controller either displays considerable function (with a low value of a particular design parameter) or is very slow (with high values of the same design parameter). In total, the control behaviour is very poor in comparison to existing fuzzy controllers, whereas NP is used effectively in the control of some other systems. Considering the concept of Control Inertia, a Sugeno-type fuzzy compensator was added to the control loop to modify the control command. A newly designed neuro-predictive control with fuzzy compensator (NPFC) improves the performance of the closed loop system significantly by the reduction of both overshoot and settling time. Furthermore, it is shown that the disturbance rejection of the NPFC controlled system as well as it parameter robustness is satisfactory.

Social Science Research Network, Jan 31, 2018

This paper reports successful development of an exact and an efficient radial basis function netw... more This paper reports successful development of an exact and an efficient radial basis function network (RBFN) model to estimate the head of gaseous petroleum fluids (GPFs) in electrical submersible pumps (ESPs). Head of GPFs in ESPs is now often estimated using empirical models. Overfitting and its consequent lack of model generality data is a potentially serious issue. In addition, available data series is fairly small, including the results of 110 experiments. All these limits were considered in RBFN design process, and highly accurate RBFNs were developed and cross validated.

Asian Journal of Control, Sep 1, 2009

In this paper, the pitch angle control of a lab model helicopter is discussed. This problem has s... more In this paper, the pitch angle control of a lab model helicopter is discussed. This problem has some specific features. As a major unusual feature, it is observed that the steady state control command is completely dependent on the setpoint, and for different setpoints, different steady state control commands are needed to keep the error around zero. Moreover, the system is one with highly oscillating dynamics. In order to solve this control problem, two controllers are designed: an artificial neural network (ANN), whose input is the setpoint, is used to provide steady state control command, and a fuzzy inference system (FIS), whose input is error, is used to provide transient control command. The total control command is the sum of the two aforementioned control commands. It is proven that both ANN and FIS are boundary-input boundary-output (BIBO) systems. Using this fact and considering two experimental assumptions, the closed-loop stability is also proven.

Engineering Letters, 2007

This paper presents a systematic approach for the design of temperature controller using genetic ... more This paper presents a systematic approach for the design of temperature controller using genetic algorithms (GAs) for thermal power plant subsystems and investigates the robustness of the designed control law. The proposed approach employs GA search for determination of the optimal PI controller parameters for a previously identified nonlinear de-superheater of a 4X 325 MW thermal power plant. Results indicate that the proposed algorithm significantly improves the performance of the thermal power plant subsystem .

International Communications in Heat and Mass Transfer, May 1, 2013

In this work system identification techniques are used to map the two-dimensional heat flux into ... more In this work system identification techniques are used to map the two-dimensional heat flux into the temperatures through a linear model supported by theoretical and numerical results. The basis of this analysis is a discrete version of the Burggraf Method saying a single component heat flux is a linear combination of the temperatures around the time of its occurrence. Taking the same approach, a linear model (i.e. a linear artificial neural network (ANN)) is employed to estimate a multicomponent heat flux as a linear function of the temperatures. A known heat flux is imposed to the direct model, then the history of heat flux-temperature data are fit to the linear mathematical model (i.e. a linear ANN) using system identification techniques. The achieved model estimates the heat flux based on a series of past and future temperatures and the estimated heat flux components are in a good agreement with the exact ones. Finally, the effect of some important factors on the results is investigated. The proposed solution to inverse heat conduction problems does not need thermophysical and geometrical parameters of the system and is robust against noises. It merely needs some series of heat flux-temperature data from solution of a reliable direct numerical model or experiment.

IEEE Sensors Journal, Oct 1, 2014

ABSTRACT If light is trapped inside a microsphere and resonance occurs, the resonance modes known... more ABSTRACT If light is trapped inside a microsphere and resonance occurs, the resonance modes known as whispering gallery modes could be employed for sensing the environment around the microsphere. The discrepancy of the resonance wavelengths for the microsphere surrounded by different media quantifies the sensing ability of the microsphere. However, the microsphere size and material are crucial factors on determining the minimum detection limit (DL) of the microsphere as a sensor. Therefore, through investigating an appropriate size and material for the microsphere, the sensing performance and efficiency of the microresonator increase. In this paper, through a comprehensive experimental study, different refractometric microspheres are presented and their optical properties are measured and analyzed. The microspheres, five different size polystyrene and one size silica microspheres, are coated with quantum dots (QDs) and the QDs are excited by an Nd:YAG laser. Then, the microspheres sensing ability is quantified when their surrounding environment is modified. According to the presented results, the microspheres’ DL, in direct proportion to the microsphere size, corresponds well to the theory. In addition, comparing the optical properties of the microspheres indicates the optimum size for the polymer microspheres to detect the environment. Furthermore, the optical properties of the silica microsphere illustrate a better performance of glass microspheres over polymer microspheres. This paper moves forward actual knowledge and evidence of extant modeling theory. It investigates a more efficient physical feature for a microsphere as a sensor. This is a key interim stage in developing more sensitive and effective sensors.

Smart Structures and Systems, May 25, 2019

IFAC Proceedings Volumes, 2008

In this paper, the pitch angle control of a laboratory model helicopter is discussed. The control... more In this paper, the pitch angle control of a laboratory model helicopter is discussed. The control has some specific features. As a main feature, it is observed that the steady state control command is completely dependent on the setpoint, so error-based controller design is not applicable to this case. Moreover, the system has a highly oscillating dynamics. In order to solve this control problem, two controllers are designed, an artificial neural network, whose input is the setpoint, is used to provide the steady state control command, and a fuzzy inference system ,whose input is the error of the system, is used to provide the transient control command. The total control command is the sum of the aforementioned two control commands. It is proved that both ANN and FIS are bounded-input boundedoutput (BIBO) systems.

ABSTRACT In this research, the fuzzy control of the yaw angle of a model helicopter is studied, p... more ABSTRACT In this research, the fuzzy control of the yaw angle of a model helicopter is studied, particularly, in order to reduce the overshoot which can be a serious problem in high inertia systems. Initially, a Sugeno-type controller is designed. This controller provides quick convergence and keeps the control input in a permitted range .Moreover, a good stability is offered by this fuzzy controller. But, a significant and repeating overshoot is observed in controlled system behaviour that is not desirable. In order to solve this problem and improve the control system, another fuzzy inference system, namely ldquofuzzy brakerdquo, is added to the closed loop circuit. Fuzzy brakepsilas task is to reduce the control input when the error is low. The proposed Sugeno-type fuzzy controller with brake (SFCB) not only vanishes the overshoot practically but also causes a considerable reduction in energy consumption, at the same time, SFCB improves the performance.

Korean Journal of Chemical Engineering, 2010

In this research, double-command control of a nonlinear chemical system is addressed. The system ... more In this research, double-command control of a nonlinear chemical system is addressed. The system is a stirred tank reactor; two flows of liquid with different concentrations enter the system through two valves and another flow exits the tank with a concentration between the two input concentrations. Fuzzy logic was employed to design a model-free double-command controller for this system in the simulation environment. In order to avoid output chattering and frequent change of control command (leading to frequent closing-opening of control valves, in practice) a damper rule is added to the fuzzy control system. A feedforward (steady state) control law is also derived from the nonlinear mathematical model of the system to be added to feedback (fuzzy) controller generating transient control command. The hybrid control system leads to a very smooth change of control input, which suits real applications. The proposed control system offers much lower error integral, control command change and processing time in comparison with neuro-predictive controllers.

Petroleum & petrochemical engineering journal, 2018

This paper critically reviews empirical models, which predict head of two-phase petroleum fluids ... more This paper critically reviews empirical models, which predict head of two-phase petroleum fluids in electrical submersible pumps. The article categorises empirical models in terms of mathematical structure and parameter identification algorithm. Categories are heuristic and artificial intelligence models. Models of the former category have fairly low accuracy and 4 or fewer parameters identifiable using non-iterative methods; conversely, models the of latter category have high accuracy and tens or even hundreds of parameters. These models require complex iterative identification algorithms. Due to availability of inexpensive digital processors, use of accurate artificial intelligence models is anticipated to broaden.