1. Introduction
Presently, to face of the depletion of primary energy resources (oil, natural gas, coal, etc.) and for the sake of preserving the environment, many countries are reviewing their energy policies and are increasingly interested in renewable energies, in particular wind energy.
To exploit this potential, various programs for the development of wind energy production have been launched. One of the main objectives of these programs is to reduce the price of the kWh produced but also the price of the investment [
1]. In fact, wind energy is only viable if its cost price is competitive. This drop in the price of the kWh produced will have to go through the improvement of the entire conversion chain. To this end, various studies have been undertaken concerning the mechanical part (blades made of composite materials, lightning of the masts, etc.), the automatic part, and diagnostics (prevention of breakdowns, mechanical regulation, etc.). In regard to the electrical part, it is articulated around two major essential axes [
2]:
Several wind energy conversion system (WECS) structures have been developed with various control techniques and are mainly: doubly feed induction generators (DFIG) and permanent magnet synchronous generators (PMSG) [
3,
4]. In the case of large–scale WECS, both DFIG and PMSG are employed and often equipped with a gearbox. Furthermore, power limitation at overrated wind speed is mainly insured by means of pitch control [
5,
6]. However, the gearbox and the pitch mechanism greatly affect the cost and robustness of the WECS and increase the maintenance requirement. In the case of small–scale WECS, a direct–drive variable–speed PMSG is widely used, and an objective can be to eliminate pitch control system to reduce the system complexity [
7,
8]. Moreover, if the WECS is equipped with full bridge back to back converters, various control techniques can be employed in order to enhance system flexibility and efficiency. Since the pitch control is eliminated, in small–scale WECS, generator operation at wind overspeed region is only obtained by flux weakening. In fact, this strategy is widely employed for motor applications [
9,
10,
11,
12,
13], but a gain interest is touched for this application in wind and marine current energy conversion systems [
7,
14,
15].
There are also other special structures, proposed in the literature for small–scale WECS with a power range of 0.5–50 kW and low speed in the range of 30 to 150 rpm, which are classified with respect to their operation as synchronous machines; these machines are called doubly salient permanent magnet (DSPM) machines [
16,
17,
18,
19,
20]. Indeed, these topologies are not new, but reemerged thanks to progress achieved in electronic and power electronic devices and magnet materials. Designated with the stator and the rotor pole numbers such as DSPM 6/4, 8/6, 12/10, and 36/24, these topologies are not convenient for low–speed applications owing to the dependence between the machine volume and the sizing torque which becomes a serious challenge.
In this context, the outer–rotor doubly salient permanent magnets generator (OR–DSPMG) is proposed for application in the small–scale kW range low speed and direct–drive WECS. Such a structure has been dimensioned in [
21,
22], for low power WT applications with a low nominal speed of 50 rpm, in order to keep the equipment costs of all the components of the machine at a low level and to be able to build the whole installation more compact and lighter. From a mechanical point of view, an outer–rotor with a low weight and inertia constitute an evident advantage because the turbine blades can be directly mounted on the machine rotor surface. The used OR−DSPMG, presented in
Figure 1, is based on the variable reluctance machine in which both the stator and the rotor are toothed to have double saliency. The stator carries small teeth (N
s). Stator poles (N
ps) also carry small teeth in the same way. This solution leads to operating the machine at low speed and producing high torque. The coils of the three phases are wound around the stator poles (concentrated winding) with each phase consisting of four coils connected in series. Rare Earth PMs (Nd-Fe-Br) are placed in the stator yoke and provide the excitation of the machine. In this structure, the rotor is placed outside and the stator inside the machine (external rotor machine). The magnetic circuit of both stator and rotor is formed by M400–50A steel type. Studied machine geometrical parameters (axial length, number of stator pole, number of stator and rotor teeth, stator and rotor yoke thickness and teeth depth, outer rotor radius, slot radius, air gap thickness, and magnet thickness) and PMs magnetization and relative permeability are given in
Table A1 (
Appendix A).
This article is the continuity of the work done in [
22], which consists of dynamic modeling and analysis of dynamic behavior and performance of the OR−DSPMG applied to a wind turbine chain at variable speeds.
The energy conversion system (10 kW−50 rpm direct–drive and grid–connected WECS) is equipped with a low–speed high torque unconventional machine (OR−DSPMG), horizontal axis wind turbine with three glass fiber blades and full bridge back to back pulse width modulation (PWM) rectifier and inverter. The system is designed to operate efficiently for all speed range, i.e., underrated and overrated wind speed, without pitch control. So, an optimal flux weakening strategy is developed to control the machine and turbine in turbine overspeed operation with constant power under given current and voltage constraint, as achieved in [
15] for a marine current energy conversion system based on PMSG. Below rated speed, maximum power point tracking (MPPT) control is released by using perturb and observe algorithm, without need to wind sensor. On the other hand, maximum torque per ampere (MTPA) control is used as a criterion for OR−DSPMG control. This strategy has a proven effectiveness for PMSG through simulation results [
7,
23,
24] and experimental validation [
7,
24]. The MTPA control is achieved by means of current regulation in d–q reference frame by moving operation point along the MTPA trajectory. Beyond the rated speed, optimal flux weakening control leads to generator operation with the maximum torque under relative current and voltage limitations. These limitations are imposed by the machine and converter and allow power limitation as described in [
14]. The corresponding d–q current references are generated and regulated. Indeed, for both MTPA and flux weakening control strategies, fuzzy logic controllers (FLC) are performed for d–q currents regulation due to these demonstrated performances (time response, steeling time, overshoot, and robustness) compared with conventional PI regulators, especially for non-linear systems [
25,
26].
Figure 1.
Outer rotor toothed doubly salient permanent magnet generator (OR−DSPMG) [
22].
Figure 1.
Outer rotor toothed doubly salient permanent magnet generator (OR−DSPMG) [
22].
The main contribution of this work is the development of MTPA and flux weakening control strategies for low–speed direct–drive WECS equipped with a new machine topology (OR−DSPMG) that can operate at underrated and overrated wind speed.
The paper is divided into five sections; the second section presents the main wind turbine and OR−DSPMG equations.
Section 3 shows the implemented control strategies: MPPT−MTPA in partial load region and flux weakening in full load region, respectively. Simulation results are presented and discussed in
Section 4; in
Section 5, conclusions are presented.
5. Conclusions
In this paper, the analysis of dynamic performance of a grid–connected WECS based direct drive OR−DSPMG is achieved. The main objective was to ensure system operation in the overspeed region with power limitation without pitch turbine blades control. Below the rated speed the MTPA strategy allows machine operation under maximum torque trajectory by mean of d–q currents regulations. As a result, a maximum average torque of 2025 Nm is obtained. Above the rated speed (4.92 rad/s), machine flux weakening control strategy was imposed in two steps: firstly, operation point was moved on current circle limit until the speed limit of this region is reached (5.44 rad/s). Secondly, flux weakening trajectory was modified in order to agree with power and voltage limitations; consequently, an overspeed of two times the rated speed is reached (9.84 rad/s). For all defined operating regions, corresponding d–q currents are calculated and regulated by means of fuzzy logic controllers, allowing to obtain satisfactory regulation performances in precision, time response, settling time, and overshoot, regardless of system settings. Similar regulation performances were obtained for DC voltage regulation, which contribute with a grid side converter control strategy to obtain an acceptable voltage and current wave form with very small THD. Furthermore, active power was transmitted to the grid with unitary power factor. However, system robustness against parameter variations is not studied in this work. An interesting future focus will be the development of an experimental platform in order to confirm the simulation results.