In this paper, efficient charging of electric vehicle battery from a considered renewable solar photovoltaic source with the help of a modified Z source with efficient boosting topology. Adapting this Z-source converter to act as a voltage gainer with a boosting function allows a solar Photovoltaic (PV) input voltage of 25VDC (Volts Direct Current) to be increased to a designed output voltage of 75VDC at a low duty ratio, resulting in minimal switching loss. The closed-loop steady-state and transient parameters at the output were analyzed and compared using modern evolutionary algorithms. The power range upheld throughout the circuit is around (300–350) W. The battery is assumed to have an impedance model of Resistor-Capacitor (RC) load with a serial range of 12 V and 7 Ah. The proposed converter achieves higher conversion efficiency by the Maximum Power Pont Tracking (MPPT) and NSGA-II/MNSGA-II (non-dominated sorting genetic algorithm) based controller algorithm for tuning the optimal design value and is validated in a MATLAB Simulink platform. In this work, we analyze closed-loop systems under the mentioned power range. The MPPT with an algorithm-based controller tends to trigger the switch in the closed-loop system to get the optimized output. The Maximum Power Point (MPP) technique implemented is an incremental conductance method for extracting solar PV power and improving load performance. Consequently, the proposed evolutionary optimization algorithm steady-state ripple factor response of the proposed MNSGA-II has a lower output side, thus achieving around 98% of the controller implementation efficiency.

In today’s world, renewable energy plays a crucial role in electrical energy conversion systems. The research is towards achieving a highly efficient power conversion system while integrating with renewable energy sources. The various boost, buck, and buck-boost topologies were involved in this research analyzed with different control methodologies to prove its technical uniqueness. Most of the literature defines eliminating technical parameters like ripple current and voltage at the parasites present in the system by implementing various multi-objective evolutionary algorithm-based controllers. Thus, this could improve the reliability and efficiency of the system.

There are various evolutionary algorithms like Particle Swarm Optimization (PSO), Genetic Algorithms (GA), and Grey Wolf Optimizer (GWO) algorithm are used for tuning Proportional Integral Derivative (PID) controller gain parameters in the conventional boost converter for getting the better optimal solution and root mean square error (RMSE) values than the conventional PID controller for different load conditions [

In [

A non-dominated sorting genetic algorithm (NSGA-II) was implemented as a multi-objective Pareto optimal model for the metal mining process to improve the production efficiency and utilization rate [

ZSBC is a converter that will operate in high voltage gain and high reliability compared to the existing converter [

Therefore, the problems caused by partial shading such as low voltage gain and drop in power level are tedious while integrating solar PV with a power conversion system [

To eliminate those above problems, the Solar PV integrated MZSBC converter with MPPT and MNSGA-II multi-objective optimization is proposed which highlights the following objective:

The MZBC is designed to produce a high voltage gain at a small duty ratio and high frequency due to the presence of coupled inductor with reduced switching losses.

The incremental conductance MPP technique helps to improve the load side performance with derived maximum power from the PV.

The MPPT and MNSGA-II optimization are used to improve the PID controller optimal gain values, the system’s dynamic performance and reduce the integral square error, which causes the system to achieve high efficiency with low-efficiency ripples at the output side.

The PID tuned gain parameters and time domain specifications like rising time, overshoot, settling time, and steady-state error obtained from different optimization algorithms are compared with the proposed MNSGA-II algorithm to show its technical importance.

By considering the efficiency, duty ratio, and controller, a comparison is made between existing and proposed converter performance.

In general, converters are electrical devices that convert voltage from alternating current (AC) to direct current (DC). This modified Z-source converter has a unique boost-up converting system that boosts input low voltage source to the designed output voltage, it gains an outcome at minimum switching loss by having a better control technique.

_{1}, an intermediate element between the PV and converter, coupled inductor La and Lb, connected through a capacitor element C_{a} and C_{b}, Switch S_{W} for turn ON/OFF purpose and, the filter components L_{c}C_{c} respectively. The MPPT and MNSGA-II controller generates a pulse to trigger the switch Sw and provide maximum power to the load. The main moto of the Z source converter in the motor vehicle gives a high gain and it attains better performance and efficiency of 98% through evolutionary data. From this evolutionary algorithm boosts up DC low voltage to DC high voltage. Therefore it mainly focused on less switching loss using a modified Z-source DC-DC boost converter employing solar energy.

In _{w} is ON, the diode will be reverse biased it will be an open circuit and the capacitors can charge the inductors; whether the circuit is in a shoot-through zero state.

When the sum of two capacitors voltage is greater than the input voltage,

The voltage across the inductors are,

In the mode 1 operation, there are the two voltage capacitors as voltage capacitor _{Ca} and voltage capacitor _{Cb} and input voltage as _{i}. From the equation, the voltage across inductors is equal to voltage across the capacitor.

In _{w} is OFF. When the switch S_{w} is turned OFF, the voltage across it keeps increasing until the input diode. D1 is turned ON. So, the voltage across the switch is clamped. In this operation mode, the voltage across inductors La is induced across inductors Lb Therefore, capacitors C_{a} and C_{b} are charged in this operation mode. Furthermore, the capacitors C_{c} and the filter inductor Lc are charged in this mode.

_{La} is the inductor current at a

_{Ca} is the capacitor current at a

As a result, the inductor current becomes zero and maintains the following switching action; therefore, this mode can isolate both the dc source and load,

Therefore, the input current value is greater than zero.

The NSGA-II algorithm has advantages like elitism, non-dominated, ranking, and crowding distance, leading to rapid convergence to optimal solutions. The NSGA-II is developed in many kinds of literature. The modified NSGA-II is implemented to solve elitism, fast non-dominated sorting approach, and diversity along with the Pareto optimal search. However, lateral diversity is still unchanged and non-dominated solutions are distributed standardized to the Pareto front. To avoid this drawback dynamic crowding distance (DCD)-based modified non-dominated sorting genetic algorithm-II (MNSGA-II) was proposed to improve the distribution of non-dominated solutions [

The crowding distance equation is given below,

_{i} is the crowding distance

The crowding distance is the value of a certain mean distance solution of its two neighboring solutions. Where CD_{i}

Individual variance (5) of the CDs with neighbors of the ith solution gives data about the different degrees of CD in different objectives.

The non-dominated sorting approach is modified with the implementation of the DCD technique. The step-by-step procedural flow of the proposed MNSGA-II incorporating DCD is given below.

Identify the Control Variable like output voltage.

Select the parameters like the number of population, the maximum number of generations, crossover and mutation probabilities.

Generate the initial population and evaluate the objective functions.

Set the generation count i = 0.

Perform simulated binary crossover and polynomial mutation.

Perform non dominated sorting for combined parent population and offspring population.

Generate population for next-generation from combined parent and offspring population using DCD.

Perform parent selection based on tournament selection.

i < i_{max}

Hence, the diversity of MNSGA-II and the Pareto optimal front is obtained with high uniformity.

Power rating | (330–400)W |
---|---|

Module data | Sunarray-S6B3612-350 T |

Ir(irradiation) W/m^{2} |
1000 @ 25°C |

Output voltage(Voc) | 25 V DC |

Output current(Isc) | 14 A DC |

Series string | 1 |

Parallel string | 2 |

Inductor L_{a}, L_{b} |
L_{a }=_{ }45μH, L_{b }=_{ }30 μH |

Capacitor C_{a}, C_{b} |
470 μF, 3600 μF |

MZSBC output voltage | 75 V DC |

MZSBC output current | 4.67 A DC |

MOSFET in MZSBC | C_{oss }=_{ }4.93 nF, V_{GS }=_{ }10 V, I_{D }=_{ }5.6A, V_{DS }=_{ }80 V |

Duty ratio | D_{ST }=_{ }0.2 |

K_{P}, K_{i,} K_{d} |
8.2, 28.63, 0.02 |

L&C filter | L_{c }=_{ }10 μH, C_{C }=_{ }800 μF |

Battery nominal voltage | 12 V DC |

Capacity | 7 Ah |

The function to be minimized is,

Subject to _{l }<_{ }x < x_{u}

f_{1 }=_{ }ISE of desired output voltage

f_{2 }=_{ }ISE of ripple in output voltage_{l}_{u}_{p}, K_{i}, K_{d} and C}; T is the simulation period length. The lower and upper bounds of PID parameters of desired output voltage where K_{p}∈[0,15], K_{i}∈[0,100], K_{d}∈[0,1] and C = {50, 150, 250, 450, 850}.

The proposed topology is simulated in the MATLAB Simulink platform and is depicted in ^{2} irradiation and 0°C temperature, the solar panel provides 25 V DC, 14 A DC power given as input to MZSBC and is boosted to 75 V DC, 4.6 A DC with a gain value of three. The power switches are triggered with the PID controller with two different proposed algorithms, like NSGA-II and MNSGA-II, for multi-objective functions, as shown in

Algorithm type | Preset values | |
---|---|---|

Information | values | |

ZN-PID | K_{p}, K_{i}, K_{d} |
6.75, 1.5, 0 |

Population size | 120 | |

Mutation constants | 24 | |

SBX crossover constants | 2 | |

NSGA-II, MNSGA-II (Proposed) | Mutation probability | 1/n |

Cross over probability | 1 | |

No. of variables (n) | 4 | |

No. of runs | 30 | |

Maximum No. of functional evaluations | 12,000 |

The Pareto optimal solution for ISE of ripple in the output voltages of NSGA-II and MNSGA-II is depicted in _{p}, K_{i}, K_{d} are tuned by the PID controller which NSGA-II and MNSGA-II obtain to reach the multi-objective optimization. Comparative performance of existing and proposed algorithm type with MZSBC is tabulated in ^{2}.

Algorithm type | K_{p} |
K_{i} |
K_{d} |
Rise time (ms) | Steady state error (V) | Settling time (ms) | Peak overshoot (V) | Efficiency @ 1000 W/m^{2} (%) |
---|---|---|---|---|---|---|---|---|

ZN-PI | 0.5 | 4.2 | 0.002 | 0.02 | 0.4 | 0.5 | 1.2 | 85.92 |

NSGA-II | 4.92 | 18.52 | 0.04 | 0.04 | 0.06 | 0.38 | 1.06 | 95.72 |

MNSGA-II (Proposed) | 8.2 | 28.63 | 0.02 | 0.05 | 0.01 | 0.25 | 1 | 98.57 |

The constant temperature of 25°C and change in irradiation with time is depicted in

The PID controller-based MPPT and NSGA-II and MNSGA-II tuning contribution of this work is to increase the response of MZSBC with a feedback system. The MPP technique used is the incremental conductance algorithm. The input voltage 25 V is boosted to 75 V while tuned by MPPT and NSGA-II/MNSGA-II based controllers, shown in

Hence, the ripple content in the output response gets reduced in the MNSGA-II technique. Therefore, it is clearly shown that the NSGA-II method oscillates between 75.3 and 74.7 obtaining ripples of 0.6 V whereas, the MNSGA-II method oscillates between 75.035 and 74.99 obtaining 0.045 V ripples. The dynamic performance analysis of solar PV integrated MZSBC of the output voltage 75 V is created in input voltage by changing the magnitude from 25 to 45 V which is created at

The step responses of ZN-PID, NSGA-II, and robust are shown in

S. no | V_{ref} (V) |
_{in} |
_{in} |
_{out} |
_{out} |
η (%) |
---|---|---|---|---|---|---|

1 | 65 | 25 | 14 | 65 | 4.92 | 91.37 |

2 | 70 | 25 | 14 | 70 | 4.8 | 96 |

3 | 75 | 25 | 14 | 75 | 4.6 | 98.57 |

Parameter | Boost | SEPIC | Proposed methodology |
---|---|---|---|

Controller | PSO | GA | MNSGA-II |

Efficiency | 94.62% | 95% | 98.57% |

The experimental result section includes the hardware setup of the proposed MNSGA-II-based MZSBC converter system with the simulation dynamic reponse and ripple in the output voltage and output current waveform using the MNSGA-II method.

FPGA Controller is established to trigger pulses for the power semiconductor switch and the generated pulse waves for the proposed optimization algorithm are depicted in

The dynamic performance study while changing input voltage and the output responses are shown in

This paper concentrates on an efficient way of charging an electric vehicle battery with a good battery and equivalent impedance models. A boosting topology is employed to charge a battery better and act as a charge controller. The power range of 350 W is utilized to charge a battery of serial 12 V, 7 Ah pack. The MZSBC converter gives a high gain from 25 V DC to 75 V DC with a small duty ratio that provides higher efficiency at the boosting level. The closed-loop simulation with different performance analyses is achieved. The PID controller gain parameters are tuned with an MPPT and NSGA-II/MNSGA-II algorithm, improving the system’s dynamic and steady-state performance. Also, the ripple content is reduced on the output side. The efficiency obtained is around 98%. A comparative study is made between proposed and existing boost converter topologies considering the efficiency and algorithm type. The simulation and hardware implementation with the best results were obtained. The dynamic performance with change in input voltage is analyzed and the reference voltage variation analysis is made to find the optimal efficiency of the proposed converter.

The authors with a deep sense of gratitude would thank the supervisor for his guidance and constant support rendered during this research.