Active/Reactive Power Control of Three Phase Grid Connected Current Source Boost Inverter Using Particle Swarm Optimization

The future is for sustainable energy (SE) sources [1, 2]. However, to utilize SE sources efficiently one needs to integrate them into the AC Power-Grid. To do so, interface electronic circuits are needed to connect SE sources to local loads or directly to the utility grid. These interfaces are known as the SE-Grid inverters. SE-Grid inverters convert direct current (DC) to alternating current (AC). They are designed with the ability to synchronize their outputs with the utility line and have as little harmonic content as possible. Inverters come in several different topologies: voltage source inverters (VSIs) [3], current source boost inverters (CSBIs) [4–6], multi-level inverters [7, 8], and matrix converters [9, 10]. VSIs have replaced CSBIs in many industrial applications [11]. However, unlike other inverters CSBIs are able to invert and boost current in a single stage and transferring power from low a DC source, for example, photovoltaic and fuel cells, to a larger AC voltage [12]. This makes CSBI one of the best choices for SE-Grid conversion systems. Here we will consider the control of active/reactive power in CSBIs with the switching pattern and control circuit which has been proposed in [13–15]. The proposed CSBI provides more robust inverter to control the voltage and the injected power. Different approaches to control active and reactive power were proposed: sliding mode control [16], fuzzy logic-based method [17] and predictive control [18]. All these control methods minimize the instantaneous errors in active and reactive power by controlling the input voltage. However, these methods are susceptible to variation of power line inductance and suffer from changing switching frequencies. In [19] and [20], direct power control technique was improved to have a constant switching frequency. In [21], the space vector modulation has been applied to a direct power controlled inverter for the purpose of having a simpler control system with fewer harmonics. One of the application to control active power in the future power system is Plug-In Electric Vehicle (PIEV) which can serve as flexible load [29]. PID control is commonly used in industrial control systems (e.g. SCADA systems and RTU’s) [22, 23]. PID minimizes the error which is the difference between a measured and a desired signal, by adjusting the process control inputs [24]. The PID controller involves three parameters: a term proportional to the error, another to the integral of error and the third to the derivative of error. The parameters of the PID controller have to be found through a design process involving the optimization of a set of desired requirements like settling time, rise time etc. Particle swarm optimization (PSO) has been developed by Kennedy and Eberhart [25]. PSO is an optimization algorithm that finds optimal solution to a problem by iteratively improving candidate solutions based on a measure of quality. It has many different applications [26–28]. PSO starts with a population of particles (candidate solutions) and by changing the particles ‘locations’ and ‘momenta’ in the search-space it explores better solutions of the problem. The particles’ locations and momenta are updated based on its local relations to other particles and by the objective function that determines the quality of solution. The process of updating the particles locations and momenta moves the swarms of particles to the optimal solution. Particles Swarm Optimization (PSO) technique is a powerful and simple method that can be used to find the PID controller parameters as to optimize design requirements. PSO-PID has been used successfully in control of automatic voltage regulators [28] and control of the linear brushless DC motor [27]. A. Sargolzaei (*) M. Jamei K. Yen A.I. Sarwat Department of Electrical and Computer Engineering, Florida International University, Florida, USA