مجله کنترل
سال دوازدهم شماره 3 (پاییز 1397)

Nowadays, the management of power consumption has been one of the main concerns for researchers, because of energy generation problems, high cost and environment pollutions. An integrated, fast and accurate system can earn satisfaction of occupants and also cost reduction. Some loads are very vulnerable to natural disasters, thus this interruptions are costly for customers. The resiliency concept is introduced accordingly. The model presented in this research is a Nanogrid, equipped with DGs, batteries and electric vehicle, and power consumer. This research devotes an effort to optimize household energy cost, occupants comfort and Nanogrid resiliency with optimal scheduling of all elements considering uncertainty of distributed energy. If needed, utility can reduce potential peaks occurrence in grid, too. Finally, the performance of the method is discussed by simulation on a typical model. Simulation results under different pricing and weather modes, show effectiveness of proposed model on reducing the quantities of each considered cost functions.

In this paper an online optimal distributed algorithm is introduced for multi-agent systems synchronization under unknown dynamics based on approximate dynamic programming and neural networks. Every agent has employed an actor-critic structure to learn its distributed optimal policy and the unknown dynamics of every agent is identified by employing a neural network approximator. The unknown dynamics are identified based on the experience replay technique where the recorded data and current data are used to adopt the approximators weights. The introduced algorithm learns the solution of coupled Hamilton-Jacobi equations under unknown dynamics in an online fashion. While the weights of the identifiers and actor-critic approximators are being tuned, the boundedness of the closed loop system signals are assured using Lyapunov theory. The effectiveness of the proposed algorithm is shown through the simulation results.

Supervisory control synthesis in discrete-event systems may encounter increasing the state cardinality. Increase in the number of states causes the computational complexity in supervisor synthesis and makes the implementation of the supervisor in industrial systems difficult. Localization of a supervisor is a method to reduce the number of states in the monolithic supervisor w.r.t. each component of the plant. Also, the synchronization of local controllers with the plant is control equivalent to the monolithic supervisor. In this paper, localization of a monolithic supervisor w.r.t. each controllable event is proposed, in order to facilitate implementation of local controllers. Two methods are proposed based on generalizing existing methods. The first method localizes a supervisor based on self-looping some states by disabled events which cannot be disabled by the corresponding local controller. The second one executes the supervisor localization based on removing transitions in the plant model that are disabled in some states of the monolithic supervisor, and are not supposed to be disabled by the corresponding local controller. In both methods, the supervisor is reduced w.r.t. the (reduced) plant model. The proposed methods are more flexible and may lead to less number of states, comparing to the results of existing method. It is proved that the two methods yield same results.

In this Study, according to microfluidic structure and bio inspired techniques, a material detection sensor has been designed and fabricated based on micro fabrication method. The capacitive sensing has been designed numerically and optimized by a familiar multi physics software. In this type of sensor, a defined aptamer has been coated on the capacitor electrode surface and by flowing the fluid through the microchannel, the targeted material would be involved between the aptamer string structure and it makes a change in capacitor value. In this research, a novel readout system based on limit cycle has been introduced and implemented. In this method, a virtual nonlinear dynamic has been supposed beside the real system and makes the whole system behave in a limit cycle (LC). By measuring LC frequency and amplitude, capacitor value is estimated.

Due to the growing demand to achieve more secure and reliable systems, development of models, analysis and design of appropriate procedures seems to be necessary. The aim of this paper is designing a controller in order to detect cyber intrusion. In this paper fuzzy neural first order hybrid Petri net is used to design a controller that is capable of detecting cyber intrusions accurately as soon as possible. The stability of the proposed intrusion detection system has been proven for any network conditions and input parameters. To evaluate controller performance, DARPA standard data set is used. The simulation results confirm proper detection rate, low of false positive rate, and also surprisingly high convergence speed.

On the grid power, the generation power should be used immediately to maintain the grid frequency constant. Therefore, grid generation power at any moment must be equal to the grid consumption power. Increasing or decreasing grid consumption power leads to the request for an increase or decrease in grid generation power; and if this request is not provided by the units in the grid generation power, a power difference will be created causing the grid frequency deviation from its nominal value. This grid power control process is called frequency control. In Iran, the main component of frequency control is provided by gas turbines. As gas turbines have been more installed in the country, the response of gas turbines to grid frequency variations has become more and more important. In order to study the frequency response of these systems, an appropriate gas turbine model is required. In this paper, the frequency model of a gas turbine has been identified. After modeling, the model's response against the errors applied to the gas turbine from the grid power is investigated by MATLAB software.