الفهرس Only 14 pages are availabe for public view
 |  | from | 180 |  |  |
| | | from | 180 | | |
Abstract
In modern power systems. the use of distributed generation systems (DG) has tremendously increased to provide power in remote and rural areas making use of the renewable energy sources such as wind energy conversion systems (WECS) , hydraulic energy and photovoltaic (PV) systems. In all of these systems pulse width modulated (PWM) inverters/rectifiers are used either to interface these systems to the main power network in case of grid connected systems or in controlling the characteristics of the electrical power generated in case of isolated or islanded systems. PWM converters are also used in flexible ac transmission systems (FACTS). This thesis proposes new control techniques to improve the performance of PWM inverters and rectifiers in modern power system applications. The thesis proposes a new control method for PWM converter used in voltage and frequency control of stand alone self excited induction generators. The new method suggests controlling only the modulation index without using a DC side chopper. The thesis also introduces a method for designing the DC side dump resistor in order to control the maximum and minimum allowable DC voltage. This reduces the system cost and reduces the capacitor current rating. The method is tested experimentally. 1’he thesis also proposes a method for utilizing the new variable DC link voltage technique for active power sharing between parallel connected induction generators by applying the frequency droop method without measuring active power. An adaptive fuzzy inverse controller is introduced to control decentralized static synchronous compensator (STATCOM) based stabilizers using only information about local generators’ speeds without knowledge of power system network data, generators parameters, regulators parameters or operating points. The identification is done online. Inverse controllers arc used to stabilize the speed deviation of the generators and the DC voltage of the SFAICOM by controlling the modulation index and the phase angle difference. The control algorithm is tested in a multimachine power system including both synchronous and induction generators with large number of fault and clearing scenarios. A nonlinear adaptive fuzzy inverse controller is proposed for control of PWM inverters with unknown circuit and filter parameters feeding unknown nonlinear loads using only output voltage sensing. The method is useful for obtaining the sinusoidal reference voltage at the load terminals in spite of the disturbance caused by the nonlinear load current and the unknown circuit and harmonic filter parameters. The method is also used for controlling the output current in case nigrid connected inverters to track a reference sinusoidal trajectory using only output current sensing in spite of the disturbance caused by the unmeasured grid voltage and unknown circuit and filter parameters. The method is tested when the unknown grid voltage is harmonic-polluted. A new method of on-line training fuzzy neural networks using particle swarm optimization (PSO) is proposed. The PSO based estimator does not suffer from the problem of estimator wind-up and is capable of finding the optimal parameters in highly complex and nonlinear search spaces. A recursive version of the objective function is introduced to make the algorithm suitable for on-line estimation and reduce the computational burden. The thesis also introduces a method for controlling PWM rectifiers by means of nonlinear adaptive fuzzy backstepping technique with adaptive estimation of the average DC load current. The method is tested under large variation of DC load current caused by starting of an induction motor. The DC load current is adaptively estimated using fuzzy basis functions which eliminate the need for a current sensor and filter to find its average. The adaptation law for DC current estimation is selected to ensure system stability.