Vector control of a grid-connected rectifier/inverter using an artificial neural network

Abstract

This is the accepted version of the paper.This version of the publication may differ from the final published version. This paper investigates how to mitigate such problems using a neural network to control a grid-connected rectifier/inverter. The neural network implements a dynamic programming (DP) algorithm and is trained using backpropagation through time. Permanent repository linkThe performance of the DP-based neural controller is studied for typical vector control conditions and compared with conventional vector control methods. The paper also investigates how varying grid and power converter system parameters may affect the performance and stability of the neural control system. Future research issues regarding the control of grid-connected converters using DP-based neural networks are analyzed. Index Terms -grid-connected rectifier/inverter, decoupled vector control, renewable energy conversion systems, neural controller, dynamic programming, backpropagation through time I. INTRODUCTIONN renewable and electric power system applications, a three-phase grid-connected dc/ac voltage-source PWM converter is usually employed to interface between the dc and ac systems. Typical converter configurations containing the grid-connected converter (GCC) include: 1) a dc/dc/ac converter for solar, battery and fuel cell applications [1,2], 2) a dc/ac converter for STATCOM applications [3,4], and 3) an ac/dc/ac converter for wind power and HVDC applications [4][5][6][7][8]. Figure 1 demonstrates the grid-connected dc/ac converter used in a microgrid to connect distributed energy resources. Conventionally, this type of converters is controlled using the standard decoupled d-q vector control approach [5][6][7][8].Notwithstanding its merits, recent studies indicate that the conventional vector control strategy is inherently limited [9,10], particularly when facing uncertainties [11]. For instance,

An adaptive recurrent neural-network controller using a stabilization matrix and predictive inputs to solve a tracking problem under disturbances

“…Each experiment started with a different initial neural-weights randomization in the range [−0.1, 0.1]. The neural controllers obtained by the best results in the first row of Table 1 replicate the neural controller performance described by Li et al (2012), which can outperform PI and ACD methods in tracking ability (as we will show in the experiments of Section 5.1). As can be seen in the table, the results are not as good as when the stabilization-matrix method, described in the next section, is used.…”

“…Limitations of these methods are that they can have slow response times to changing reference commands, can take considerable time to settle down from oscillating around the target reference state (Dannehl et al, 2009), and have difficulty recovering from short-circuit faults in either the generator or the power-grid. Hence neural-network based solutions have been proposed to overcome these difficulties, in this control problem and related ones (Qiao et al, 2008b(Qiao et al, , 2009aLi et al, 2012;Venayagamoorthy et al, 2002;Park et al, 2004;Qiao et al, 2008a;Venayagamoorthy et al, 2003).…”

“…Recent studies show how a single action network can be trained with BPTT to control a GCC under fixed plant behavior (Li et al, 2012). However, for real-life applications, the plant behavior can change; system parameters can exhibit random variations; voltages coming into the system from the power grid can fluctuate; short circuits can occur.…”

“…However, for real-life applications, the plant behavior can change; system parameters can exhibit random variations; voltages coming into the system from the power grid can fluctuate; short circuits can occur. Hence the action network needs to become more adaptive than demonstrated by Li et al (2012). Adaptive behavior can be enabled by modifying the action network to have neural-context units which respond to the changing behavior of the plant, thus making the action network into a RNN.…”

“…to currents i d and i q , and voltages v d and v q . In this simpler d-q reference frame, the voltages and currents evolve according to the following 2-dimensional vector differential equation (see Li et al (2011Li et al ( , 2012 for further derivation):…”