1 Introduction
1.1 Background and context
1.2 Literature survey
References  Year  Location  Structure  Evaluation Metrics  Methodologies  

E  V  T  R  HOMER  MA  HOMER vs. MA  
[28]  2022  Bangladesh  PV/grid  √  √  √  –  √  –  – 
[29]  2020  Egypt  PV/WT/DsGl/BESS  √  √  √  –  √  –  – 
[30]  2022  Greece  PV/WT/grid  √  √  √  –  √  –  – 
[31]  2021  Egypt  PV/WT/DslG/BESS/FC  √  √  √  –  √  –  – 
[32]  2021  Canada  PV/WT/DslG/BESS/TPH  √  √  √  –  √  –  – 
[41]  2021  Egypt  PV/WT/BESS  √  –  √  –  √  GWO, PSO, WHO, GA  √ 
[50]  2021  Mexico  PV/DslG/BESS  √  √  √  –  √  NSGA–II  √ 
[51]  2022  China  PV/WT/FC  √  √  √  √  √  VIKOR  √ 
[43]  2021  Egypt  PV/WT/DslG/BESS  √  √  –  √  √  PSO, GA  √ 
[44]  2021  India  PV/WT/BESS  √  –  √  –  √  BSA  √ 
[45]  2020  Algeria  PV/WT/DslG/BESS  √  √  √  √  √  PSO  √ 
[52]  2021  France  PV/WT/BESS  √  –  –  √  √  PSO, GA  √ 
[53]  2020  Ecuador  PV/DslG/BESS  √  –  √  √  –  PSO–BPSO  – 
[54]  2021  Iran  PV/WT/BESS  √  –  –  –  –  TS & HS  – 
[55]  2018  KSA  PV/WT/DslG/BESS  √  √  √  –  –  MOSaDE  – 
[56]  2022  –  PV/WT/BESS  √  √  √  –  NSGA–II  –  
[57]  2022  India  PV/WT/DslG/BESS  √  √  √  –  √  PSO, SSA  √ 
[58]  2022  –  PV/WT/DslG/BESS  √  √  √  –  √  HPSODE–FSM  √ 
Current  2022  Egypt  PV/WT/DslG/BESS  √  √  √  √  √  AVOA, GOA, GPC  √ 
1.3 Research gaps and contributions

Developing a robust mathematical model for an autonomous solar/wind/diesel/battery/converter HRES to power the 24h load demand of a remote urban community in Marsa Matruh city, Egypt, considering actual load and renewable resources data. Meanwhile, presenting an adequate energy management strategy is suggested to coordinate the power flow between various energy sources in RESs which are fully exploited.

Proposing a new application of the AVOA optimization algorithm to determine the optimal configuration and components’ capacities of the HRES under study. The objective function is formulated as multiple objectives to minimize the total net present cost and CO_{2} emissions while maintaining the system’s loss of power supply reliability at the lowest level.

Validating and comparing the performance of the AVOA with HOMER, the trusted global standard software in hybrid power system modeling, and uptodate metaheuristic methods of the grasshopper optimization algorithm (GOA) and the Giza pyramid construction (GPC).

Providing a systemic and comprehensive energy–economic–environmental analysis of the winning HRES design to understand better the system behavior with the proposed solution based on AVOA.
1.4 Paper organization
2 System description and mathematical modeling
3 Formulation of the design optimization problem
3.1 Design criteria
3.1.1 Total net present cost
Item  Value 

Number of iterations (MaxIT)  100 
Population size (nPop)  50 
Number of dimensions (dim)  4 
Minimum and maximum values (l_{b}, u_{b})  Eq. (37) 
3.1.2 Penalty of emissions
3.1.3 Cost of energy
3.2 Design constraints
3.2.1 Capacity constraints
3.2.2 Battery lifetime constraints
3.2.3 Diesel generator operational constraints
3.2.4 System reliability constraints
3.3 Objective function
4 Proposed solution method
4.1 The African vultures optimization algorithm
4.1.1 Finding the best vulture in any group
4.1.2 Defining the rate of vultures’ starvation
4.1.3 Exploration phase
4.1.4 Exploitation phase
4.1.4.1 First phase
4.1.4.2 Second phase
4.2 Development of AVOA for the optimal HRES design
5 Case study
Input  Value  Constraint  Value (%) 

Nominal discount rate  13.25%  Maximum capacity shortage/LPSP  0 
Expected inflation rate  4.8%  Minimum renewable fraction  0 
Project lifetime  25 years  Operating reserve as % of load  5 
System fixed capital cost  0 $  Operating reserve as % of PV output  5 
System fixed O&M cost  0 $/yr  Operating reserve as % of WT output  5 
Capacity shortage penalty  0 $/kWh  Battery maximum SOC  100 
Carbon dioxide penalty  30 $/ton  Battery minimum SOC  40 
6 Results and discussion
6.1 Design optimization results
Metric  Optimizer  

AVOA  GOA  GPC  
Min  346,614.05  346,685.17  347,222.81 
Max  348,073.10  382,924.71  357,824.08 
Mean  346,789.82  352,921.34  350,800.81 
Median  346,685.17  348,935.70  350,772.36 
Std. deviation  319.40  9,995.73  2,484.44 
Variance  102,015.78  99,914,630.11  6,172,418.11 
PV  WT  DslG  BESS  CON  ObjFn  Execution time  Rank  

kW  Qty  kW  Qty  kW  $  Seconds  
AVOA  42  0  27  36  32  346,614  18.66  1 
GOA  62  0  22  87  38  362,064  19.51  3 
GPC  46  0  31  39  34  353,253  22.76  2 
HOMER  43  6  15  104  26.2  370,881  130.6  4 
6.2 Economic analysis
Cost component ($)  PnCE ($)  TNPC ($)  LCOE ($/kWh)  

CapC  O&MC  RepC  SavC  
AVOA  99,800  147,681  78,268  4334.1  25,198.4  346,614  0.0947 
GOA  134,050  127,159  82,173  1199.6  19,880.9  362,064  0.0990 
GPC  110,450  146,133  76,308  4372.9  24,734.6  353,253  0.0966 
HOMER  167,859  84,070  94,520  11,402.6  35,833.5  370,881  0.239 
Element  Method  CapC  O&MC  RepC  SavC 

PV  AVOA  42,000  4460.5  0  0 
GOA  62,000  6584.6  0  0  
GPC  46,000  4885.3  0  0  
HOMER  42,995  4565.1  0  0  
WT  AVOA  –  –  –  – 
GOA  –  –  –  –  
GPC  –  –  –  –  
HOMER  60,000  3185.2  11,451.3  5828.1  
DslG  AVOA  32,400  136,000  60,991  3719.6 
GOA  26,400  107,299  45,338  469.934  
GPC  37,200  133,495  57,674  3720.1  
HOMER  18,000  83,114  36,839  640.38  
BESS  AVOA  12,600  3823.3  13,276  0 
GOA  30,450  9239.7  32,084  0  
GPC  13,650  4141.9  14,382  0  
HOMER  36,400  11,042  42,960  4,432  
Conv  AVOA  12,800  3398.5  4001.6  614.42 
GOA  15,200  4035.7  4751.9  729.62  
GPC  13,600  3610.9  4251.7  652.82  
HOMER  10,462  2777.2  3269.5  501.88 
6.3 Energy analysis
6.4 Emission analysis
Diesel generator data  CO_{2} emissions  

Capacity (kW)  Hours  Fuel (L/year)  Lifetime (years)  (kg/year)  
AVOA  27  3626  29,958  4.13  79,089.1 
GOA  22  3511  23,636  4.27  62,399.1 
GPC  31  3100  29,406  4.83  77,633.4 
HOMER  15  4022  18,251  3.73  47,778 
7 Conclusions and perspectives

The proposed AVOA algorithm achieved superior results concerning the objective function value compared to other approaches. It achieved a minimum TNPC and PnCE of 346,614$ and COE (0.0947 $/kWh), equivalent to 6.5 and 60.4% savings compared to HOMER results, respectively.

The design based on AVOA efficiently served the load demand with zero LPSP with an acceptable value for a renewable fraction of 40.38%.

The metaheuristic algorithms showed fast execution time, with AVOA being the first ranked with an average of 18.66 s, followed by GOA (19.51 s) and GPC (7.84 s). In contrast, HOMER has taken significantly longer than the metaheuristic algorithms to find the optimal solution (130 s), which is timeconsuming.