Power Management Control of an Autonomous Photovoltaic/Wind Turbine/Battery System

Abstract

The study presents an optimal control approach for managing a hybrid Photovoltaic/Wind Turbine/Battery system in an isolated area. The system includes multiple energy sources connected to a DC bus through DC/DC converters for maximum power point tracking. The proposed hybrid MPPT approach (HMPPT) manages the energy production from different sources, while the power flow method is used to balance the load and renewable power. The study shows that integrating the HMPPT algorithm and power flow approach results in improved system performance, including increased power generation and reduced stress on the batteries. The study also proposes an accurate sizing method to further improve system efficiency. The study demonstrates the effectiveness of the proposed approach by presenting results for twelve different days with varying weather conditions. The results show that the proposed approach effectively manages the energy production and load, resulting in optimal system performance. This study provides valuable insights into the optimal control of hybrid renewable energy systems, and highlights the importance of considering different energy sources and optimal sizing for maximizing system efficiency.

1. Introduction

Wind and photovoltaic energies are clean and environmentally friendly. Nevertheless, their power output is intermittent, posing a threat to the electrical power system resilience. Hence, the introduction of storage is a feasible solution to balance the demand and supply of electricity by storing extra energy and producing it when needed. As a result, a hybrid system is created that is safer and more environmentally friendly [1,2,3,4].

When the wind speed changes or solar irradiation deteriorates, power point tracking approaches are used to extract the optimal load power. A wide variety of MPPTs are used to track the maximum power point of photovoltaic panels [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24] and wind turbines [25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40]. While they all aim to boost power, each performs differently from the others.

Because of its simplicity, the Perturb and Observe (P&O) method is the most widely employed to determine the MPP point for PV systems and wind turbine (WTb) generators, but this technique suffers from steady-state oscillations [5,6,7,8]. Other methods have been employed to evaluate quick and effective MPPT strategies for PV systems, such as MPPT algorithms based on voltage and current [9,10]. Regarding smart MPPT methods [11,12], they are more frequently used to address the non-linear characteristics of solar PV panels. A table on a microcomputer is also used in [13] to track MPP. In [14,15], mathematical equations or numerical approximations utilizing the curve-fitting approach are used to characterize the non-linear behavior of the PV generator. In the short-circuit method [16], the voltage of the PV generator at the MPP is roughly linearly related to its opencircuit. These techniques appear to be easy and costeffective, but they are unable to adapt to shifting climatic conditions.

The P&O (Perturb and Observe) technique is commonly used in control schemes due to its simplicity. However, it has a drawback of oscillation, which cannot be completelyeliminated [17,18,19,20]. The conductance incremental approach described in [21] needs a complicated control circuit. Additionally, intelligent methods based on control by MPPT were introduced [22,23,24]. The fuzzy logic controller (FLC) optimizes the magnitude of the increment to obtain fast and fine tracking. This method is widely used because of its advantages, but the controller depends on speed and power variations. The adaptive fuzzy logic controller (AFLC) method is also widely used for its advantages of fast response.

In wind turbines, the P&O algorithm [25,26,27,28,29,30,31,32] is the most used, and other techniques such as Optimal Torque Control (OTC) [25], Tip Speed Ratio (TSR), Power Signal Feedback (PSF), Fuzzy and Adaptive Logic Controller (FLC and AFLC), Genetic Algorithm (GA), Adaptive Neuro-Fuzzy Inference System (ANFIS), Artificial Neural Network (ANN)-based controller and Particle Swarm Optimization (PSO) are also frequently used [33,34,35,36,37,38,39]. Due to its simplicity of implementation, the hill climb search (HCS) algorithm is the most widely used strategy. It compares the power that was previously given with that following the disturbance. The Power Signal Feedback (PSF) approach generates a reference power signal to achieve the best power output. On the other hand, the Optimal Torque Control (OTC) approach adjusts the generator torque to an optimal level for different wind speeds to maximize power output [36]. Power and speed variations are the AFLC’s input controllers, and the reference speed variation is the output. The rules will be dependent on variations in speed and power to converge to the optimum point.

Power and speed variations are the input controllers for the AFLC, with the reference speed variation serving as the output. The rules for convergence to the optimum point are dependent on variations in speed and power. In multi-source renewable energy systems (MSRES), energy management control (EMC) is mandatory. EMC is used in a wide range of applications, from renewable energy sources like photovoltaics (PV) and wind with their simple or multi-storage, to automotive traction using batteries, fuel cells (FC) and supercapacitors (SC). Control strategies for EMC have been investigated in numerous publications [40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71].

In this work, an optimal control of a hybrid photovoltaic/wind/battery system is implemented and validated through Matlab/Simulink. The suggested sizing strategy makes use of the total incident annual monthly average [1]. This method enables estimation of the size of the two renewable generators. Many PV, wind turbine, and battery parameters have been determined experimentally and used in numerous simulations to generate more precise models. The proposed optimization strategy is associated with the P&O and AFLC strategy, and is based on a hybrid maximum power point tracking (MPPT) method that includes the advantages of each strategy alone. The hybrid MPPT method offers several advantages, including higher power, efficiency and response time, which results in reduced battery stress. Measured solar irradiances and wind speeds during twelve different days corresponding to each month in the year 2021 are used, along with a chosen load profile to supply power to a residential home.

To manage the different powers, a power flow or supervision method is used. It is worth noting that this technique has been used in previous works [51,56,65], but not yet on a hybrid PV/wind turbine/battery system. The method is simple, easily implemented and does not require heavy computations. The load power and hybrid renewable power are always compared, and when there is a lack of power, the batteries, if charged, supply the load alone or in combination with the other sources; if there is an excess of power, it charges the batteries. The key accomplishments of this work are a substantial increase in renewable power and reduced battery stress in a multi-source (PV/Wind turbine) system. The hybrid MPPT (HMPPT) method is associated with the proposed management method, and the accuracy of the sizing method used contributes to the feasibility of the system. Simulation results using Matlab/Simulink show that the system can handle varying solar irradiance and wind speed conditions.

2. Design and System Description

A permanent magnet synchronous generator (PMSG)-equipped wind turbine system, DC/DC and DC/AC photovoltaic converters, storage batteries, solar irradiance sensors and wind speed sensors make up the studied system (Figure 1). The power maximization in each generator is performed using the proposed hybrid MPPT algorithm (HMPPT), and a power flow strategy is used to control the different sources.

The implemented design method is based on the total incident energy’s annual monthly average. Monthly energy of each generator and load demand are calculated to find the different areas of panels and wind turbine [1]. The different areas of the PV and wind turbine generator are given by Equations (1) and (2) [1]:

$${\mathrm{S}}_{\mathrm{pv}}=\max \left(\frac{{\mathrm{E}}_{\mathrm{Load},\mathrm{m}}}{{\mathrm{E}}_{\mathrm{pv},\mathrm{m}}}\right)$$

(1)

$${\mathrm{S}}_{\mathrm{wind}}=\max \left(\frac{{\mathrm{E}}_{\mathrm{Load},\mathrm{m}}}{{\mathrm{E}}_{\mathrm{wind},\mathrm{m}}}\right)$$

(2)

The PV, wind and load energy produced are given by Equation (3):

$${\mathrm{E}}_{\mathrm{pv}}={\mathsf{\eta }}_{\mathrm{pv}}.{\mathrm{S}}_{\mathrm{pv}}.{\mathrm{E}}_{\mathrm{irr}}$$

(3)

$${\mathrm{E}}_{\mathrm{wind}}={\mathrm{P}}_{\mathrm{wind}}.\Delta \mathrm{t}=\left(1/2\right).\mathsf{\rho }.\mathrm{S}.{\mathrm{V}}_{\mathrm{wind}}^3.{\mathrm{C}}_{\mathrm{p}}$$

(4)

$${\mathrm{E}}_{\mathrm{Load}}={\mathrm{S}}_{\mathrm{pv}}.{\mathrm{E}}_{\mathrm{pv}}+{\mathrm{S}}_{\mathrm{wind}}.{\mathrm{E}}_{\mathrm{wind}}$$

(5)

with:

$${\mathsf{\eta }}_{\mathrm{pv}}={\mathsf{\eta }}_{\mathrm{pv}-\mathrm{STC}}.\left[1-{\mathsf{\beta }}_{\mathrm{oc}}.\left({\mathrm{T}}_{\mathrm{j}}-{\mathrm{T}}_{\mathrm{j}-\mathrm{STC}}\right)\right]$$

(6)

$$\{\begin{array}{c}{\mathrm{S}}_{\mathrm{pv}}={\mathrm{k}}_{\mathrm{perc}}\left({\mathrm{E}}_{\mathrm{Load}}/{\mathrm{E}}_{\mathrm{pv}}\right)\\ {\mathrm{S}}_{\mathrm{wind}}=\left(1-{\mathrm{k}}_{\mathrm{perc}}\right)\left({\mathrm{E}}_{\mathrm{Load}}/{\mathrm{E}}_{\mathrm{wind}}\right)\end{array}$$

(7)

The monthly energies produced are

$$\{\begin{array}{c}{\mathrm{E}}_{\mathrm{pv},\mathrm{m}}=\left({\displaystyle \sum }_{\mathrm{m}=1}^{12}{\mathrm{E}}_{\mathrm{pv}}\right)/12\\ {\mathrm{E}}_{\mathrm{wind},\mathrm{m}}=\left({\displaystyle \sum }_{\mathrm{m}=1}^{12}{\mathrm{E}}_{\mathrm{wind}}\right)/12\\ {\mathrm{E}}_{\mathrm{Load},\mathrm{m}}=\left({\displaystyle \sum }_{\mathrm{m}=1}^{12}{\mathrm{E}}_{\mathrm{Load}}\right)/12\end{array}$$

(8)

where k_perc is the PV source’s fraction of the load, and (1 − k_perc) is the wind source’s fraction of the load.

As a result, it is obtained:

$$\{\begin{array}{c}{\mathrm{S}}_{\mathrm{pv}}={\mathrm{k}}_{\mathrm{perc}}\left({\mathrm{E}}_{\mathrm{Load},\mathrm{ave}}/{\mathrm{E}}_{\mathrm{pv},\mathrm{ave}}\right)\\ {\mathrm{S}}_{\mathrm{wind}}=\left(1-{\mathrm{k}}_{\mathrm{perc}}\right)\left({\mathrm{E}}_{\mathrm{Load},\mathrm{ave}}/{\mathrm{E}}_{\mathrm{wind},\mathrm{ave}}\right)\end{array}$$

(9)

The following equations determine the number of PV panels and wind turbines:

$$\{\begin{array}{c}{\mathrm{S}}_{\mathrm{pv},\mathrm{final}}={\mathrm{N}}_{\mathrm{pv}}.{\mathrm{S}}_{\mathrm{pv},\mathrm{unit}}\\ {\mathrm{S}}_{\mathrm{wind},\mathrm{final}}={\mathrm{N}}_{\mathrm{wind}}.{\mathrm{S}}_{\mathrm{wind},\mathrm{unit}}\end{array}$$

(10)

The average consumed energy is given by:

$${\mathrm{E}}_{\mathrm{load}-\mathrm{ave}}={\mathrm{E}}_{\mathrm{pv},\mathrm{ave}}.{\mathrm{S}}_{\mathrm{pv},\mathrm{unit}}+{\mathrm{E}}_{\mathrm{wind},\mathrm{ave}}.{\mathrm{S}}_{\mathrm{wind},\mathrm{unit}}$$

(11)

In

Table 1. PV and wind system configurations.

Months	Eir (kWh/m2)	Ta (°C)	Tj (°C)	ηpv	Vwind (m/s)	Epv (kWh/m2)	Ewind (kWh/m2)	ELoad (kWh)
January	180.14	16.00	21.63	0.1130	6.50	15.90	191.47	664.82
February	190.11	14.00	19.94	0.1138	6.00	16.89	136.03	477.15
March	200.45	18.00	24.26	0.1118	5.00	17.51	87.15	311.52
April	292.32	19.00	28.14	0.1101	4.70	25.14	70.05	260.00
May	200.03	25.00	31.25	0.1087	4.10	16.99	48.05	178.13
June	260.41	28.00	36.14	0.1065	4.20	21.67	49.99	188.78
July	290.70	31.00	40.08	0.1048	3.50	23.79	29.89	122.29
August	290.80	36.00	45.09	0.1025	3.70	23.29	35.32	140.30
September	185.56	30.00	35.80	0.1067	3.90	15.46	40.02	149.51
October	190.10	22.00	27.94	0.1102	5.50	16.36	116.00	408.60
November	190.10	18.00	23.94	0.1120	6.60	16.63	193.98	673.97
December	160.26	14.00	19.01	0.1142	6.80	14.29	219.23	757.78
						Epv,ave = 18.66	Ewind,ave = 101.43	ELoad,ave = 361.07

, various calculations are summarized to demonstrate the different configurations of the solar system’s linkage with the wind power system. It can be observed that the average photovoltaic energy is about 19.80 kWh/day, and the average wind energy is about 134.12 kWh/m². Since the average load energy required is 476.86 kWh, the combination of 38 solar panels and 0 wind turbines (

Table 2. The number of panels and wind turbines calculated.

${\mathbf{k}}_{\mathbf{perc}}$	Spv (m2)	Swind (m2)	Npv	Nwind	Spvfinal (m2)	Swind,final (m2)	ELmean (kWh)
0.00	0.00	3.56	0.00	2.00	0.00	6.90	699.87
0.10	1.93	3.20	3.00	1.00	2.60	3.45	398.52
0.20	3.87	2.85	5.00	1.00	4.34	3.45	430.92
0.30	5.80	2.49	7.00	1.00	6.08	3.45	463.31
0.40	7.74	2.14	9.00	1.00	7.81	3.45	495.71
0.50	9.67	1.78	12.00	1.00	10.42	3.45	544.30
0.60	11.61	1.42	14.00	1.00	12.15	3.45	576.69
0.60	11.61	1.42	14.00	1.00	12.15	3.45	576.69
0.70	13.54	1.07	16.00	1.00	13.89	3.45	609.08
0.80	15.48	0.71	18.00	1.00	15.62	3.45	641.48
0.90	17.41	0.36	21.00	1.00	18.23	3.45	690.07
1.00	19.35	0.00	23.00	0.00	19.96	0.00	372.53

) corresponds to the required load energy (486.02 kWh). However, since the studied system is a hybrid system (PV/wind turbine), only the configuration of 8 panels and 1 wind turbine is closest to the required load energy (523.47 kWh).The calculation of battery capacity can be written as [32]:

$${\mathrm{C}}_{\mathrm{Batt}}=\left({\mathrm{d}}_{\mathrm{aut}}.{\mathrm{E}}_{\mathrm{load},\mathrm{m}}\right)/\left({\mathrm{U}}_{\mathrm{Batt}}.\mathrm{PDP}.{\mathsf{\eta }}_{\mathrm{Batt}}.{\mathrm{N}}_{\mathrm{m}}\right)$$

(12)

The batteries number can be calculated as

$${\mathrm{N}}_{\mathrm{batt}}=\mathrm{ENT}\left[{\mathrm{C}}_{\mathrm{batt}}/{\mathrm{C}}_{\mathrm{batt}-\mathrm{u}}\right]$$

(13)

with: ${\mathrm{d}}_{\mathrm{aut}}$ as the days of autonomy (days), ${\mathrm{E}}_{\mathrm{load},\mathrm{m}}$ the monthly load consumed (kWh/day) N_m: 31 days,${\mathrm{U}}_{\mathrm{batt}}$ as the voltage battery (V), $\mathrm{PDP}$ as the depth of discharge, ${\mathsf{\eta }}_{\mathrm{batt}}$ the efficiency battery and ${\mathrm{C}}_{\mathrm{batt}-\mathrm{u}}$ as the chosen battery capacity.

Table 3. Component numbers.

PV Panels	Wind Turbine	Batteries
03	01	02

summarizes all the number components to be used. Finally, the total maximum power of Photovoltaic panels is P_pv-tot = 3 × 110 = 640 W_p, while the maximum wind turbine power is about 01 × 900 W = 900 W and 02 batteries of (12 V, 100 Ah) are used.

3. Optimization Methods

To optimize the strategy in PV and wind turbine generators, two MPPT approaches (P&O and AFLC) are adopted.

3.1. P&O Method

This algorithm principle is described in Figure 2, and explained in the flowchart (Figure 2) [1].

3.2. Adaptative Fuzzy Logic Controller (AFLC)

Adaptive Fuzzy Logic Controller (AFLC) is an upgraded version of the Fuzzy Logic Controller (FLC) that includes an adaptive mechanism for tuning the controller parameters in realtime based on changes in the system and environmental parameters (Figure 3). The goal of the AFLC is to improve the performance of the FLC by adjusting its duty-cycle or other control parameters to optimize the system response. The AFLC typically includes two main components: a fuzzy logic controller and a learning mechanism. The fuzzy logic controller is similar to the FLC, and is used to process inputs from sensors and generate control signals. The learning mechanism is responsible for monitoring the system performance and environmental parameters and adjusting the FLC parameters accordingly. In the AFLC, the inputs to the fuzzy logic controller typically include the PV module’s voltage and current, which are added to the preceding values to produce the average value. This input is then processed by the fuzzy logic controller, which generates an output signal that is used to control the system. The learning mechanism is used to modify the FLC duty-cycle or other parameters based on the performance of the system and environmental parameters. This mechanism typically includes a set of rules that are used to adjust the FLC parameters based on feedback from sensors and other sources.The AFLC is a powerful tool for optimizing the performance of renewable energy systems by providing real-time control and adaptation to changes in the system and environmental parameters. By incorporating fuzzy logic control and adaptive learning mechanisms, the AFLC can improve the efficiency, reliability and stability of renewable energy systems.

The controller MAMDANI type is shown in

Table 4. AFLC rules.

Error (e)	Variation of Error (Ce)
	NB	NM	NS	ZE	PS	PM	PM
NB	NB	NB	NM	ZE	ZE	ZE	ZE
NM	NB	NM	NM	ZE	NM	PS	PS
NS	NB	NB	NB	NB	PM	PS	PM
ZE	NB	NB	NS	ZE	PS	PM	PB
PS	NM	NS	ZE	PS	PM	PB	PB
PM	NS	PB	PB	PB	PB	PB	PB
PB	ZE	PB	PB	PB	PB	PB	PB

Where: NB, NS, NM, ZE, PM, and PB refer to linguistic terms used to describe the degree of membership. NB: “Negative Big” represents a high degree of negative membership, NS: “Negative Small” represents a moderate degree of negative membership, NM: “Negative Medium” represents a moderate degree of negative membership, ZE: “Zero”represents a neutral degree of membership, PM: “Positive Medium” represents a moderate degree of positive membership, PS: “Positive Small” represents a moderate degree of positive membership and PB: “Positive Big”—represents a high degree of positive membership.

, along with functions for membership in seven classes [1]. By using linguistic terms to describe the input and output variables, the FLC can make control decisions based on qualitative descriptions rather than precise numerical values.

3.3. Hybrid MPPT Approach

A hybrid MPPT approach is suggested (HMPPT). This is a combinationof P&O and AFLC methods. First, the optimal values are calculated (Equation (14)), and then the proposed algorithm gives the chosen best power values (Equation (15)).

$$\{\begin{array}{c}{\mathrm{P}}_{\mathrm{PV}-\mathrm{opt}}=\{{\mathrm{P}}_{\mathrm{PV}-\mathrm{P}\&\mathrm{o}},{\mathrm{P}}_{\mathrm{PV}-\mathrm{AFLC}}\}\\ {\mathrm{P}}_{\mathrm{Tb}-\mathrm{opt}}=\{{\mathrm{P}}_{\mathrm{Tb}-\mathrm{P}\&\mathrm{o}},{\mathrm{P}}_{\mathrm{Tb}-\mathrm{AFLC}}\}\end{array}$$

(14)

$$\{\begin{array}{c}{\mathrm{P}}_{\mathrm{PV}-\mathrm{best}}=\max \left({\mathrm{P}}_{\mathrm{PV}-\mathrm{opt}}\right)\\ {\mathrm{P}}_{\mathrm{Tb}-\mathrm{best}}=\max \left({\mathrm{P}}_{\mathrm{Tb}-\mathrm{opt}}\right)\end{array}$$

(15)

where: ${\mathrm{P}}_{\mathrm{PV}-\mathrm{opt}}$ and ${\mathrm{P}}_{\mathrm{Tb}-\mathrm{opt}}$ are the different PV and wind turbine power values of each MPPT method (P&o and AFLC),respectively, and ${\mathrm{P}}_{\mathrm{PV}-\mathrm{best}}$ and ${\mathrm{P}}_{\mathrm{Tb}-\mathrm{best}}$ are the selected best optimal PV and wind turbine power chosen by the HMPPT algorithm, respectively.

An application ismade during a whole day to show the performance of the HMPPT method (Figure 4).

The three MPPT methods (P&O, AFLC and HMPPT) have been applied. It is evident from Figure 4 that the suggested HMMPT approach reacts more quickly than the P&O and AFLC methods.

4. Simulation Study

The power of each generator is maximized using the proposed HMPPT strategy (Figure 5).The hybrid HMPPT MPPT approach is a combination of the P&O and AFLC methods.

In a wind turbine generator, field-oriented control (FOC) is used for DC bus regulation to keep constant the voltage DC bus constant whatever the wind speed variations. This is used to control the currents and voltages of the stator windings of the generator, which are then converted to DC voltage by a rectifier and fed into the DC bus.

4.1. Measurements ofSolar Irradiation, Temperature and Wind Speed Profiles

A measurement acquisition equipment was used in the lab to detect the sun radiation, temperature and wind speed (Figure 6). It is essentially composed of sensors in order to transfer the different signals to a data processing interface, and then to a PC where they will be displayed using ACQUIsol software in realtime.

4.2. Simulation under Measured Profiles of Solar Irradiation, Temperature and Wind Speed

Matlab/Simulink is used to run the simulations, which consider the measured solar irradiance and wind speeds profile for each day (Figure 7).

The chosen load profile is as shown in Figure 8. In our work, we use a hybrid HMPPT. The power obtained from the PV generator under the three MPPTs is shown in Figure 9. Three zoomed-in views of the photovoltaic power during three different profile days are displayed in Figure 9a–c, corresponding respectively to medium, low and high solar irradiance.

The wind power obtained using the HMPPT strategy is presented in Figure 10. Three zoomed-in views of the wind turbine power during three different profile days are displayed in Figure 10a–c, corresponding respectively to high, medium, and low wind speed velocity.

The renewable power is defined as shown in Figure 11. Three zoomed-in views of the renewable power during three different profile days are presented in Figure 11a–c, corresponding respectively to medium, low and high solar irradiance and wind speed velocity.

$${\mathrm{P}}_{\mathrm{R}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{w}}={\mathrm{P}}_{\mathrm{p}\mathrm{v}}+{\mathrm{P}}_{\mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d}}$$

(16)

It can be said that when using the hybrid method HMPPT and for each energy source, significant power improvements can be realized when compared to using a single MPPT technique. The required power by the batteries is reduced, as shown in Figure 12. Three zoomed-in views of the battery power during three different profile days are displayed in Figure 12a–c.

The battery state of charge (SOC) is a critical parameter that must be evaluated to ensure safe charging and discharging procedures. Estimating the SOC helps to protect the battery from overcharging or deep discharging, thereby extending its life. The battery SOC is maintained within the following limits

$${\mathrm{SOC}}_{\min }\le \mathrm{SOC}\le {\mathrm{SOC}}_{\max }$$

(17)

where SOC_max is the maximum SOC value and SOC_min is the minimum SOC value.

For battery control, we usually use a Buck-Boost converter for the charging and discharging process (Figure 13).

In our case, the SOC was controlled between 30% and 90% (Figure 14). Three zooms of the SOC during three different profile days, respectively, are shown in Figure 14a–c, corresponding respectively to medium, low and high solar irradiance and wind speed velocity It is noticed that the SOC in maintaining between the SOC_min and the SOC_max whatever the variations of solar irradiance and wind speed during the three different profile days. It is maintained to its maximum value of 90 percent whatever the day profile with some variations to reach a SOC_min of 56.57% in the first day (Figure 14a), 48.98% in the second day (Figure 14b) and 61.98 % in the third day. The batteries are less stressed and discharge less, which is very important for their life span.

It can be said that on any given day, with its own weather variables, the suggested hybrid technique produces the best SOC results. In comparison to the other methods, it maintains its maximum value of 90 percent. With the proposed hybrid method, the batteries are less stressed and discharge less, which is very important for their lifespan. Furthermore, it is noted in Figure 15 that the battery voltage stays close to the reference voltage of 24V, which is likely the desired voltage for the system. The hybrid MPPT approach (HMPPT) produces the best values in terms of battery voltage, whatever the day chosen. This one quickly reaches its value of 24V and is kept longer compared to the other methods. Three zoomed-in views of the voltage battery during three different profile days, respectively, are shown in Figure 15a–c.

The battery voltage is maintained to its maximum value of 24V whatever the day profile with some variations to decrease to 23.77 V in the first day (Figure 15a), 23.61 V in the second day (Figure 14b) and 23.75 V in the third day. The batteries have not discharged too much, which is very important for their life span.

The findings suggest that the HMPPT approach is effective at optimizing the power output of the renewable energy source and regulating the battery voltage to ensure the safe and optimal operation of the system.

5. Power Control of the Studied System

5.1. Literature Review on Energy Management Control

Energy management control (EMC) is a crucial component in multi-source renewable energy systems (MSRES). Its applications range from renewable energy sources such as PV and wind, with their simple or multi-storage systems, to automotive traction using batteries, fuel cells (FC) and supercapacitors (SC). Numerous research articles [40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65] have investigated various control strategies for these systems, as outlined in

Table 5. Literature review of some systems under study with their components with EMC.

Year	System under Study	Components							References
		PV	WTb	Batteries	Diesel Generator	Fuel Cells	Hydropower	SC
2009	Autonomous system	X	X	X	X	X	X		[40]
2010	Autonomous system	X		X				X	[41]
2012	Micro-grids	X	X	X					[42]
2013	Domestic micro-grids			X		X			[43]
2013	Micro-grids	X	X						[44]
2014	Traction motor			X		X		X	[45]
2014	Autonomous network	X		X		X			[46]
2014	Electric vehicle	X		X		X			[47]
2014	Autonomous system	X				X			[48]
2015	Electric car	X		X		X			[49]
2016	Water pumping system	X	X	X					[50]
2017	Autonomous system	X	X	X	X				[51]
2017	Electric vehicle			X				X	[52]
2018	Autonomous system	X		X					[53]
2018	Hybrid vehicle			X			X		[54]
2018	Water pumping system	X	X	X	X				[55]
2019	Autonomous system	X	X	X	X				[56]
2020	Micro-grids	X	X	X					[57]
2020	Standalone system	X		X		X			[58]
2020	Grid-PV system	X		X		X			[59]
2021	Micro-grid systems	X	X	X	X	X	X	X	[60]
2021	Isolated renewable energy system			X				X	[61]
2021	Isolated hybrid micro-grids	X		X	X				[62]
2021	Micro-grid systems	X	X						[63]
2021	Micro-grids	X	X	X	X	X	X	X	[64]
2021	DC micro-grids	X		X					[65]
2022	Micro-grids	X	X	X					[66]
2022	Micro-grids	X	X	X	X	X			[67]
2022	Stand-alone system		X	X					[68]
2022	Stand-alone system	X	X	X					[69]
2022	Electric vehicle			X		X		X	[70,71]

. While the technologies used in each study differ, most applications focus on isolated systems for electrification [40,45,46,47,48,49,50,51], microgrids [42,58,60,62,63,64] and multi-storage in traction and electric vehicles [47,50,52]. In most works, optimal component sizing has been studied [3,4,5,6,7,8,9,10,11,12,13], and it has been concluded that efficient control management strategies are necessary for correctly sizing the different components of MSRES. Some of these strategies are based on linear controllers [51,52,53], while others employ more intelligent methods [47,60,61,62,63,64,65]. Most of the works focus on simulation, implementation, economic study, optimization and analysis of the system’s performance and environmental impact. The mainfocus of EMC methods is controlling the power from different sources to supply the load and protect the storage system. Some EMC methods use “if-else” statements in the decision algorithms [56], while others employ more intelligent and predictive methods [60,61,62,63,64,65,66,67,68,69,70,71]. The following table provides a summary of the literature review of the systems studied.

5.2. Control of the Studied System

The studied system PV/Wind turbine with batteries with the HMPPT strategy has been controlled using a proposed power management control (Figure 16). There is always a comparison between load power and hybrid renewable power. When there is a lack of power, the batteries, if fully charged, can supply the load alone or in compensation with other sources; it also charges batteries if there is an excess of power. For this, four switches are used in the suggested power flow system (Figure 17).

The system uses four switches (K₁, K₂, K₃, K₄) to control the flow of power between the different sources and the load. When there is sufficient power from the renewable sources to meet the load demand, switch K₁ is closed to allow the power to flow directly to the load. At the same time, switch K₄ is closed to allow any excess power to be stored in the batteries. If the renewable sources are not producing enough power to meet the load demand, switch K₁ is opened and switch K₂ is closed to allow the batteries to supply power to the load. The batteries can either supply power alone or in conjunction with any remaining power from the renewable sources. If the batteries become depleted, switch K₃ is closed to allow an alternate power source, such as a generator or grid power, to supply power to the load. At the same time, switch K₄ is closed to allow the batteries to be charged from the alternate power source. The different switches operate as shown in (

Table 6. Different states of the switches.

Switch States				Powers	SOC
K1 = 1	K2 = 0	K3 = 0	K4 = 0	Phyb = PLoad	SOC
K1 = 1	K2 = 0	K3 = 0	K4 = 1	Phyb > PLoad	SOC > SOCmax
K1 = 1	K2 = 1	K3 = 0	K4 = 0	Phyb > PLoad,	SOC < SOCmax
K1 = 1	K2 = 0	K3 = 1	K4 = 0	Phyb < PLoad,	SOC > SOCmin
K1 = 0	K2 = 1	K3 = 0	K4 = 0	Phyb < PLoad	SOC < SOCmin
K1 = 0	K2 = 0	K3 = 0	K4 = 1	PLoad = 0, Phyb > 0	SOC ≥ SOCmax

).

The load profile depends on PV, wind and battery powers:

$${\mathrm{P}}_{\mathrm{Load}}={\mathrm{P}}_{\mathrm{PV}}+{\mathrm{P}}_{\mathrm{wind}}\pm {\mathrm{P}}_{\mathrm{Batt}}.$$

(18)

Figure 18 shows the various switch signals, while Figure 19 depicts the various powers. It can be concluded that the load power was satisfied during the twelve typical days throughout the year due to good sizing and the effective management of the different power sources, in accordance with the power balance equation.

The batteries’ capacity was used up to 44.93%on the first day (Figure 19a), 62.96% on the second (Figure 19b), and only 32.36% on the third (Figure 19c). As expected, the power discharge represented only a small quantity (indicated by the negative parts in a red color) when compared to the results obtained using the proposed hybrid method, which placed less stress on the battery.

6. Conclusions

The paper presents the optimal control of a hybrid photovoltaic/wind turbine/battery system, and analyzes and compares the results from different findings. There was a significant improvement in power for each energy source when using the hybrid HMPPT method, as compared to using a single MPPT technique. The required power from the batteries has been greatly reduced. The main contribution of this paper is the reduction of stress on the storage batteries in a multi-source system through the combination of an accurate sizing approach and a hybrid MPPT algorithm with a power flow method. The results demonstrate that the different energy sources were managed optimally to meet the load demand, even under varying weather conditions. The findings confirm the effectiveness and viability of the suggested control method for the entire year.