Intelligent Power Management Control for Hybrid Wind Solar Battery Systems Connected to Micro-Grids

The growing development of renewable energies, notably photovoltaic and wind power, brings undeniable advantages both economically and environmentally, thus promoting the transition towards cleaner and more sustainable energy. However, the isolated use of these sources significantly limits their benefits and performance [1]. It is in this context that the hybridization of renewable sources, grouped within a multi-source system called a hybrid system, emerges as a solution [2]. This approach aims to ensure a stable and uninterrupted power supply, meeting the required energy needs.

Hybrid systems [3], which combine renewable sources such as photovoltaic generators (PVG), wind turbines (WT) with batteries as energy storage systems (ESS), and are connected to the electrical distribution grid [4], are specifically designed to provide the necessary power to the load, regardless of meteorological fluctuations and variations in the availability of individual sources [5]. Furthermore, the common use of an energy storage system (ESS) is observed in standalone applications [6]. Renewable energy hybrid systems require various control techniques to ensure efficient energy transfer, requiring considerable technical attention and generating research in this field [7]. Simulations have been conducted on a hybrid energy system to evaluate state of charge (SOC) control under different load demand scenarios and weather conditions [8]. In parallel, solar and wind are identified as clean and renewable energy sources. Photovoltaic systems face challenges due to the nonlinear properties of solar radiation, while the variability of wind speed makes obtaining constant energy from wind power complex [9]. The integration of artificial intelligence techniques in energy harvesting is being explored to improve performance, quality, and speed compared to conventional approaches [10]. Additionally, PV-wind-battery systems with multifunctional control techniques are being presented. These concepts examine energy flow in bidirectional converters of micro grids, taking into account the battery state of charge [11].

The objective of this contribution is to design and develop a micro grid combining wind energy, solar energy, and energy storage systems, with a focus on the use of renewable energy sources [12]. An energy management system is proposed to maintain power balance in the micro grid, providing configurable and flexible control for various load demand variations and energy consumption scenarios [13], as well as fluctuations in renewable energy sources [14].

The remainder of the document is structured as follows: Sections 2-3 elaborate on the overall system design, converter topologies, and control techniques. Section 4 introduces the system management algorithm. The obtained results are then discussed in Section 5. Finally, the document concludes with Section 6.

3.1 Solar energy conversion system

Photovoltaic cells, fundamental components of solar technology used in the production of solar panels and photovoltaic generators, convert light into electrical energy through the photovoltaic effect. They are typically arranged in series and parallel to form a solar panel. However, in specialized documentation, it is common to adopt a simplified electrical model, comprising a single diode (as illustrated in Figure 2), to represent a silicon photovoltaic cell and simulate its behavior. This model consists of a photocurrent source, a nonlinear diode, a series resistance reflecting internal losses, and a resistance connected in parallel with the diode to account for leakage current to the ground.

2.png

Figure 2. Electrical circuit equivalent of a GPV cell

From this circuit, these equations can be derived [15, 16]:

$I_{p v}=I_{p h}-I_d-I_{s h}$                     (1)

The current flowing through these components is determined by the voltage applied across them.

$V_d=V_{p v}+I_{p v} R_S$               (2)

According to the Shockley diode equation, the current flowing through the diode can be expressed as:

$I_d=I_0\left[e^{\left(\frac{q V_d}{n k T}\right)}-1\right]$                 (3)

According to Ohm's Law, the current passing through the shunt resistor can be determined by dividing the voltage across the resistor by its resistance, as follows [17]:

$I_{s h}=\frac{V_d}{R_{s h}}$                (4)

$I_{p v}=I_{p h}-I_0\left[e^{\left(\frac{q\left(V_{p v}+I_{p v} R_s\right)}{n k T}\right)}-1\right]-\frac{V_{p v}+I_{p v} R_s}{R_{s h}}$         (5)

where, I_ph: photocurrent; I_pv: cell output current; V_pv: cell’s output voltage; V_d: voltage across diode; I₀: reverse saturation current; R_sh: shunt resistance; R_s: serie resistance; Ns: number of cells in one module; n: diode ideality factor; k: Boltzmann’s constant (1.381 × 10^-23); T: absolute temperature equals-273.15℃ and q: electron charge (1.602 × 10^-19C).

Two solar panels are equipped with an adjustable irradiance constant by the user, with a default value set to 1000 W/m².

The described model was implemented using Matlab/Simulink software, taking into account the provided parameters outlined in Table 1.

Table 1. Parameters of the GPV

Next, the system underwent testing across four levels of solar irradiation: 1000W/m² (standard), 800W/m², 600W/m², and 400W/m², with the temperature held constant at 25℃.

Figure 3 illustrates the current and power characteristic curves plotted against the voltage of the PV array, represented respectively by I=f(V) and P=f(V).

According to the characteristics P=f(V) illustrated in Figure 3, it is observed that the maximum power that the photovoltaic generator can produce depends on the level of irradiation. Therefore, to ensure optimal operation at maximum power, the implementation of a control system is essential to monitor the maximum power point (MPP). Table 2 provides additional information on this subject.

3.png

Figure 3. I=f(V) and P=f(V) of GPV for different irradiances

Table 2. Ppv_Max under different irradiations

3.2 Wind energy conversion system

The Wind Energy Conversion System (WECS) consists of a wind turbine, a Permanent Magnet Synchronous Generator (PMSG), a DC-DC boost converter, and a controller. The extraction of power from the wind follows the description provided in reference [18]. Identify applicable funding agency here. If none, delete this text box.

$\mathrm{P}_{\mathrm{w}}=1 / 2 \rho \operatorname{Av} 3 \mathrm{C}_{\mathrm{p}}(\lambda, \theta)$                 (6)

Here, ρ represents the air density in kg/m³, A denotes the area swept by the rotor blades in m², and v stands for the wind velocity in m/s. C_p, the power coefficient, is determined by the tip speed ratio (TSR) and pitch angle, represented by λ and θ, respectively, and are designed accordingly [19] .

The different characteristics of the wind turbine’s output power as a function of its speed for various wind speeds are shown in Figure 4 and the parametres of the wind system used in this work is shown in Table 3.

4.png

Figure 4. Wind turbine mechanical characteristics across varying wind speeds

Table 3. Parameters of WECS

3.3 Modeling of PMSG

The widely employed mathematical model of the Permanent Magnet Synchronous Generator (PMSG) in the rotor reference dq frame is depicted as follows:

$\left\{\begin{array}{l}V_{s d}=R_g i_d-\omega_e \varphi_{s q}+\frac{d \varphi_{s d}}{d t} \\ V_{s q}=R_g i_q-\omega_e \varphi_{s d}+\frac{d \varphi_{s q}}{d t}\end{array}\right.$               (7)

where, V_sd and V_sqdenote the stator voltage, i_dand i_q signify the stator current, φ_sdand φ_sqdenote the stator flux in frames [20]. R_g represents the stator resistance, while ω_e denotes the electrical rotation speed of the PMSG, defined as follows [21]:

$\omega_e=p_r \omega_{r m}$                 (8)

where, p_r: is the number of pole pairs.

The stator flux expressions are as follows:

$\left\{\begin{array}{l}\varphi_{s d}=L_d i_d+\varphi_f \\ \varphi_{s q}=L_q i_q\end{array}\right.$             (9)

where, L_dand L_q represent the generator inductances on the d and q axes, respectively, and φ_f the permanent magnetic flux.

The expression for the electromagnetic torque is as follows:

$T_e=\frac{3}{2} p_r\left(\varphi_{s d} i_q-\varphi_{s q} i_d\right)$                (10)

By substituting Eq. (7) into Eq. (9), the model of the Permanent Magnet Synchronous Generator (PMSG) in the synchronous reference frame can be expressed as follows:

$\left\{\begin{array}{l}V_{s d}=R_g i_d+L_d \frac{d i_d}{d t}-L_q i_q \omega_e \\ V_{s q}=R_g i_q+L_q \frac{d i_q}{d t}+L_d i_d \omega_e+\omega_e \varphi_f\end{array}\right.$                     (11)

The electromagnetic torque in Eq. (10) is rewritten as follows:

$T_e=\frac{3}{2} p_r\left[\varphi_f i_q+\left(L_d-L_q\right) i_q i_d\right]$                     (12)

Assuming equal inductances on the d and q axes for the PMSG (L_d / L_q), the expression for the electromagnetic torque in Eq. (12) can be described as follows [22]:

$T_e=\frac{3}{2} p_r i_q \varphi f$                (13)

The parameters system is shown in Table 4.

Table 4. Parameters of filter and inverter

3.4 Modeling of battery system

The choice of solar battery type is crucial due to its use in multi-source systems. They are categorized into three distinct types based on their technology: maintenance-free batteries, with an efficiency of around 80%; lead-acid batteries, with an efficiency of around 90%; and lithium-ion batteries, the most efficient, with an efficiency of around 98% [23-25]. The use of lithium-ion batteries in multi-source energy management systems is justified by several key factors that meet the specific requirements of these systems. Compared to lead-acid or nickel-cadmium batteries, lithium-ion batteries offer a higher energy density, making them suitable for energy storage applications where long-term reliability is necessary. They also exhibit low energy loss during storage and discharge processes and can respond quickly to charging and discharging needs. Furthermore, lithium-ion batteries are generally lighter and more compact than other types of batteries.

Modeling a lithium-ion battery is intricate, encompassing the consideration of numerous parameters including battery capacity, internal resistance, voltage, temperature, and charge/discharge rate. Mathematical models such as the equivalent circuit model are frequently employed to characterize these batteries and forecast their performance under varying usage scenarios. In this study, Figure 5 illustrates the equivalent electrical circuit of the battery utilized, while Eqs. (14) and (15) represent the battery voltage (V_Bat) and state of charge (SOC) [26, 27].

5.png

Figure 5. RC model of the battery

$V_{\text {Bat }}=E_0-R_{\text {Bat }} I_{\text {Bat }}-k \cdot \int \frac{I_{\text {Bat }}}{Q} d t$         (14)

$S O C=100\left(1+\frac{\int I_{\text {bat }} d t}{Q}\right)$              (15)

where, Q is the battery capacity and I_Bat denotes the battery current, which operates in: charging and discharging modes, depending on the power generated by solar and wind energy [28]. The battery parameters are defined in Table 5.

Table 5. Parameters of battery

The battery also operates in these two modes based on the energy constraints dictated by the state of charge (SOC) limits, as determined by [29].

$\mathrm{SOC}_{\min } \leqslant \mathrm{SOC} \leqslant \mathrm{SOC}_{\max }$                 (16)

Figure 8 presents the profiles of wind speed and solar irradiance used for simulation tests aimed at analyzing and evaluating the performance of the management system adopted in various environmental scenarios.

8.png

Figure 8. Profiles of wind speed and solar irradiance levels

We subjected the system to two different profiles and numerically simulated its behavior for two load scenarios. In the first case, the load power is kept constant, while in the second case, it varies. The analysis of the system's behavior under these conditions is presented below:

Case 1. Fixed load power

Figures 9 and 10 illustrate, respectively, the evolution of the load power (Pload) and that produced by the renewable energy sources (RES), notably those from the photovoltaic generator (Ppv) and the wind turbine (Pw) and the evolution of state of charge of battery for constant load.

For such an operating scenario, there are primarily four (4) distinct time intervals during which the batteries alternate between two (2) intervals in charging mode and the other two in discharging mode, as explained below. So, for each time interval, we observe that:

From 0 to 1s: The temporal analysis of the power (see Figure 9) shows that the power levels evolve rapidly (response time of approximately 0.001s) to reach the levels corresponding to the weather conditions (irradiation and wind speed) shown in Figures 3 and 4 within this time frame. The power from the renewable energy sources (RES) is 4500 W, which exceeds the load demand of 3500 W. This excess power is used to charge the battery, increasing the SOC to 50.0007%, as shown in Figure 10.

From 1s to 2s: During this time interval, the power demanded by the load remains at 3500 W, while the sum of the powers produced by the renewable energy sources (RES) is approximately 3400 W. Therefore, to meet the load power demand, the missing power of 100 W is provided by the batteries, resulting in a decrease in SOC from 50.0007% to 50.0006%, as shown in Figure 10 during the period from 1 second to 2 seconds.

From 2s to 3s: The sum of the powers produced by the RES is 2900 W, which is less than the load demand of 3500 W. Therefore, the batteries continue to discharge, (battery discharge mode, "Decrease in SOC to 50.0006%") approximately 49. 998% to provide the missing power 600 W.

From 3s to 4s: For this last time interval, it is observed from Figure 9 that the sum of the powers produced by the RES is equal to 4100 W, which is greater than the one required for the load. Consequently, the batteries transition from the discharge mode to the charge mode ("Increase in SOC from 49.998% to 49.9985%"), as shown in Figure 10.

9.png

Figure 9. Power at different locations for constant load

10.png

Figure 10. State of charge of battery for constant load

Case 2. Variable load power

In the case where the load power varies, as illustrated in Figure 11, and under the same conditions of irradiation and wind change previously considered, Figures 11 and 12 respectively depict the powers and the state of charge of the batteries (SOC).

Based on the two preceding figures, the following observations can be made for the four time intervals:

From 0 to 1s: During this time interval, the load power is 1000 W, while the sum of the powers produced by the RES is equal to 4600 W, which is sufficient to meet the load power requirements. The excess power is then stored in the batteries. This is justified by the increase in SOC observed in Figure 12.

From 1s to 2s: In this time interval, there is a decrease in irradiation level, while the wind speed and load power remain unchanged compared to the previous interval. Thus, the sum of the powers produced by the RES is approximately 3900 W, which is sufficient to meet the load demand and provide the excess to the batteries, which continue to charge as shown in Figure 12.

From 2s to 3s: During this interval, the load power has increased to 2900 W, due to the weather changes, the sum of the RES powers corresponding to this environment is 2800 W. Therefore, the lack of power to satisfy the load demand is compensated by transitioning from the battery charging mode to discharge mode, resulting in a decrease in SOC.

From 3s to 4s: The load power demanded is 3200 W, while the sum of the powers produced by the RES is 4200 W, resulting in an excess of power. The batteries are thus operating in charge mode, justified by the increase in SOC in Figure 12.

11.png

Figure 11. Power at different locations for variable load

12.png

Figure 12. State of charge of battery for variable load

However, following this analysis of the different scenarios close to reality, it is proven that the system operates correctly according to the specifications. The analysis of SOC variations is crucial for the battery's lifespan and the overall performance of the hybrid system. Indeed, proper SOC management is essential to ensure optimal performance and extend the battery's life. In terms of energy efficiency, the SOC is generally maintained within an optimal range between 20% and 80%, which we have adopted as a condition for our simulations. It is important to note that an SOC that is too high (Soc=100%) or too low (SOC=0%) can accelerate the chemical degradation of battery cells. However, based on the analysis specific to each previously discussed time interval, we observe that the SOC varies between 49.998% and 50.001% for Case 1: Fixed load power (see Figure10) and 50 % to 50, 0013 %) for Case 2: Variable load power. Consequently, it is properly maintained within the defined range, thereby prolonging the battery life. Additionally, the temporal power analysis shows , for both cases, that the response time is quick and sufficient to provide energy instantly to reach the optimal power level during fluctuations in energy demand, thus ensuring the stability of the electrical grid.