Powertrain control of a solar photovoltaic-battery powered hybrid electric vehicle

Abstract

This paper proposes a powertrain controller for a solar photovoltaic battery powered hybrid electric vehicle (HEV). The main objective of the proposed controller is to ensure better battery management, load regulation, and maximum power extraction whenever possible from the photovoltaic panels. The powertrain controller consists of two levels of controllers named lower level controllers and a high-level control algorithm. The lower level controllers are designed to perform individual tasks such as maximum power point tracking, battery charging, and load regulation. The perturb and observe based maximum power point tracking algorithm is used for extracting maximum power from solar photovoltaic panels while the battery charging controller is designed using a PI controller. A high-level control algorithm is then designed to switch between the lower level controllers based on different operating conditions such as high state of charge, low state of charge, maximum battery current, and heavy load by respecting the constraints formulated. The developed algorithm is evaluated using theoretical simulation and experimental studies. The simulation and experimental results are presented to validate the proposed technique.

Keywords

battery management system / hybrid electric vehicles (HEVs) / maximum power point tracking (MPPT) / solar photovoltaic

Introduction

The future vehicle uses alternative fuels such as electricity and fuel cells in contrast to the present internal combustion powered vehicle due to depleting nature of fossil fuels, cost, and environmental issues [¹]. Electric vehicle (EV) is one of the alternatives to internal combustion (IC) engine powered vehicle in the recent past [²]. EVs are powered by electric motors and battery as the energy storage element. It is estimated that by 2050, around 100 million of EVs will be sold per year worldwide [³]. However, the EVs are still not popular because of two important challenges, driving range and cost. The driving range (or) range of the EVs is defined as the maximum possible distance traveled by the vehicle per charge. It depends upon the power of the electric motor, the capacity of the battery, and the overall efficiency. Increasing the capacity of the battery will lead to the increase in the cost of the vehicle and hence most of the research work is involved in increasing the overall efficiency of the system.

The overall efficiency of the EV can be improved by increasing the efficiency of the individual components in the powertrain. In IC engine based vehicles, the powertrain comprises of the main components that generates power and delivers it to the road surface. This includes the engine, the transmission, the drive shafts, the suspension, and the wheels. But in EVs, the electric motors, the battery and transmission system are considered as the powertrain components. By improving the powertrain algorithm, the benefits of the powertrain can be maximized. Various control strategies are found in the literature to address the powertrain problem, namely multilevel power allocation based on empirical mode decomposition and fuzzy control [⁴], the global optimal design method [⁵], the rule-based control strategy [⁶], the selection of optimal drive-train configuration [⁷], the power split method between two sources under real-life load fluctuations [⁸], energy flow control through the battery EV and to extract maximum power from photovoltaic (PV) [⁹], and optimal powertrain design for the fuel-cell HEV [¹⁰]. Alternatively, the battery can be charged while driving using a method called regenerative braking [¹¹–¹⁴]. But using this method, only a small amount of energy can be stored in the battery, which cannot increase the driving range in a significant way. Combining two energy sources for power is an alternative idea to increase the range but it will also increase the cost of the system. To optimize the cost and energy, the solution could be installing PV panels in EVs.

In combining energy sources, the battery and PV panels are well studied techniques developed in the literature. The rechargeable battery integrated to a solar PV panel through a DC to DC converter is known as a solar PV-battery power source [¹⁵]. For PV power generation, the PV panel capacity is very large compared to the battery, and moreover batteries are used to attenuate the energy fluctuation resulted from the PV panel due to environmental conditions [¹⁶,¹⁷]. But the EV cannot be equipped with more number of PV panels due to the space availability and hence, the panel capacity is less compared to the battery capacity. In the solar PV-battery powered HEV, the solar PV and/or the battery will deliver the output power. A novel control approach for integrating a solar PV with a battery for residential and EV application is found in Ref. [¹⁸]. The converter for integrating the solar PV and battery is discussed in Refs. [¹⁹,²⁰]. The voltage regulation along with maximum power extraction for a solar PV-battery fed system for EV application is presented in Ref. [²¹]. A drive control system for a solar PV-battery powered EV is proposed in Ref. [²²]. A bidirectional power flow strategy for the gird connected PV-battery system is presented in Ref. [²³].

Battery health is one of the main features that affect the overall efficiency of the EV powertrain system. A good battery management system (BMS) can improve the life and health of the battery [²⁴]. Since there are more charging and discharging cycles in EVs, especially in plug-in EV with vehicle-to-grid (V2G) and grid-to-vehicle (G2V) [²⁵,²⁶], and due to the intermittent behavior of renewable resources, batteries will experience several incomplete and partial charge/discharge cycles, which, in turn, reduces its lifetime [¹⁶]. Besides, frequent power exchange between EV and battery will decrease the battery lifetime, which leads to reduced storage capacity [²⁷]. Thus, there is a need to ensure the safe and optimal use of the energy inside the battery, which is the task of a BMS [²⁸]. In addition, BMS also ensures appropriate intervention of the battery system when it is operated in an abnormal condition.

A PV powered system, either for PV power generation or PV hybrid EV, can be optimized if it operates at maximum power point (MPP). A PV system can operate under non-uniform working conditions such as temperature change, irradiance change, and shadowing phenomena [²⁹]. Therefore, the MPP for each panel may vary [³⁰,³¹]. Numerous maximum power point tracking (MPPT) algorithms are evolved to track the MPP, including the Lagrange interpolation based particle swarm optimization (PSO) [³¹], the deterministic cuckoo search (CS) [³²], the perturb and observe (P&O) based least power point tracking [³³], the adaptive P&O based power point tracking [³⁴], the enhanced adaptive P&O [³⁵], the P&O based gray wolf-assisted hybrid control [³⁶], the image of PV module based method [³⁷], and the fast hybrid technique [³⁸]. A comparative evaluation of various MPPT algorithms is found in Refs. [³⁹–⁴²]. The P&O is one of the most popular MPPT algorithms, due to its simplicity and ease of implementation [¹⁸,⁴³].

The main motive of this paper is to develop an algorithm to optimally utilize the battery and solar PV to power the EV such that the overall system respects BMS, load regulation and maximum power extraction from solar panels. To achieve this, a state space model of the complete system is developed to derive the transfer function model of three different subsystems to design the individual controllers for load regulation, battery management, and maximum power extraction. Then, a high-level control algorithm (HLCA) is developed to switch between individual controllers to ensure good BMS and maximum power extraction. These individual controllers are termed as lower level controllers in this paper which are MPPT control, load regulation control, and battery charge controllers, namely constant voltage charging (CVC) and constant current charging (CCC). The MPPT algorithm is designed using the P&O method which ensures PV panels to operate at MPP. The CCC and CVC controllers are designed to charge the battery by regulating constant current and constant voltage respectively. Both these controllers are designed using PI controllers. The parameters such as V_batt, i_batt and SOC are monitored continuously to implement the logic. Four different operating conditions such as low SOC, high SOC, maximum battery current, and heavy load are considered for the analysis. The performance of the lower level controllers is individually evaluated for all the above conditions by performing simulation and experimental analysis. Finally, the HLCA is implemented in real-time, which switches between these lower level controllers to respect the constraints formulated. The change in operating conditions is ascertained by varying the load. The values of the parameters are carefully chosen so as to execute all the operating conditions. The important contribution of this work are development of a high-level algorithm to optimally use solar PV and battery by respecting the constraints of BMS, verification of the high-level algorithm by performing numerical simulation, and development and verification of a solar PV-battery powered EV prototype in the laboratory.

Mathematical modeling of the solar PV-battery powered HEV

The block diagram of the solar PV-battery powered HEV is shown in Fig. 1 while the actual circuit model is demonstrated in Fig. 2. The proposed system consists of solar PV modules, two buck converters (buck converter-I and II), battery, and load. Considering Fig. 2, the dynamics of V_pv is expressed as

{\dot{V}}_{\mathrm{pv}}=k_{\mathrm{1}}(i_{\mathrm{pv}}-i_{\mathrm{L}\mathrm{1}}{u}_{\mathrm{1}}).

where

k_{\mathrm{1}}={\displaystyle \frac{\mathrm{1}}{C_{\mathrm{pv}}}}.

The dynamics of i_{\mathrm{L}\mathrm{1}} is expressed as

{\dot{i}}_{\mathrm{L}\mathrm{1}}=k_{\mathrm{2}}(V_{\mathrm{pv}}{u}_{\mathrm{1}}-V_{\mathrm{batt}}),

where

k_{\mathrm{2}}={\displaystyle \frac{\mathrm{1}}{L_{\mathrm{1}}}}.

The dynamics of the battery voltage can be expressed as

{\dot{V}}_{\mathrm{batt}}=k_{\mathrm{3}}(i_{\mathrm{L}\mathrm{1}}-i_{\mathrm{batt}}-u_{\mathrm{2}}{i}_{\mathrm{L}\mathrm{2}}),

where

k_{\mathrm{3}}={\displaystyle \frac{\mathrm{1}}{C_{\mathrm{1}}}}.

The dynamics of the battery current can be expressed as

{\dot{i}}_{\mathrm{batt}}=k_{\mathrm{4}}(i_{\mathrm{L}\mathrm{1}}-u_{\mathrm{2}}{i}_{\mathrm{L}\mathrm{2}})+k_{\mathrm{5}}(V_{\mathrm{batt}}-E)-k_{\mathrm{6}}{i}_{\mathrm{batt}},

where

k_{\mathrm{4}}={\displaystyle \frac{\mathrm{1}}{C_{\mathrm{1}}{R}_{\mathrm{2}}}}\mathrm{;}{k}_{\mathrm{5}}={\displaystyle \frac{\mathrm{1}}{C_{\mathrm{batt}}{R}_{\mathrm{1}}{R}_{\mathrm{2}}}}\mathrm{;}{k}_{\mathrm{6}}={\displaystyle \frac{\mathrm{1}}{C_{\mathrm{1}}{R}_{\mathrm{2}}}}+{\displaystyle \frac{R_{\mathrm{2}}+R_{\mathrm{1}}}{C_{\mathrm{batt}}{R}_{\mathrm{1}}{R}_{\mathrm{2}}}}.

The dynamics of battery SOC can be represented as

\mathrm{S}\dot{\mathrm{O}}\mathrm{C}=1+k_7{i}_{\mathrm{b}\mathrm{a}\mathrm{t}\mathrm{t}},

where

k_{\mathrm{7}}={\displaystyle \frac{1}{C_{\mathrm{0}}}}.

The dynamics of i_L2 is expressed as

i_{\mathrm{L}\mathrm{2}}=k_{\mathrm{8}}(V_{\mathrm{batt}}{u}_{\mathrm{2}}-V_{\mathrm{0}}),

where

k_{\mathrm{8}}={\displaystyle \frac{\mathrm{1}}{L_{\mathrm{2}}}}.

The dynamics of V_c2 is expressed as

{\dot{V}}_{\mathrm{c}\mathrm{2}}={\dot{V}}_{\mathrm{0}}=k_{\mathrm{9}}(i_{\mathrm{L}\mathrm{2}}-k_{\mathrm{10}}{V}_{\mathrm{0}}),

where

k_{\mathrm{9}}={\displaystyle \frac{\mathrm{1}}{C_{\mathrm{2}}}}\mathrm{;}{k}_{\mathrm{10}}={\displaystyle \frac{\mathrm{1}}{Z_{\mathrm{L}}}}.

Defining the state variables,

x_{\mathrm{1}}=V_{\mathrm{pv}}\mathrm{;}{x}_{\mathrm{2}}=i_{\mathrm{L}\mathrm{1}}\mathrm{;}{x}_{\mathrm{3}}=V_{\mathrm{batt}}\mathrm{;}{x}_{\mathrm{4}}=i_{\mathrm{batt}}\mathrm{;}

x_{\mathrm{5}}=\mathrm{S}\mathrm{O}\mathrm{C}\mathrm{;}{x}_{\mathrm{6}}=i_{\mathrm{L}\mathrm{2}}\mathrm{;}{x}_{\mathrm{7}}=V_{\mathrm{0}},

the state space model of the complete system can be expressed as

(1)\dot{X}=\mathit{A}\mathit{X}+\mathit{B}\mathit{u}+{\mathit{B}}_{\mathrm{1}}\mathit{u}\mathit{X}+{\mathit{B}}_{\mathrm{2}}\mathit{E},

where

\mathit{A}=\left[\begin{array}{ccccccc}\hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill -k_{\mathrm{2}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill k_{\mathrm{3}}\hfill & \hfill \mathrm{0}\hfill & \hfill -k_{\mathrm{3}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill k_{\mathrm{4}}\hfill & \hfill k_{\mathrm{5}}\hfill & \hfill -k_{\mathrm{6}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill k_{\mathrm{7}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill -k_{\mathrm{8}}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill k_{\mathrm{9}}\hfill & \hfill -k_{\mathrm{9}}{k}_{\mathrm{10}}\hfill \end{array}\right],

\mathit{B}=\left[\begin{array}{ccc}\hfill k_{\mathrm{1}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \end{array}\right],\hspace{0.17em}\mathit{u}=\left[\begin{array}{c}\hfill i_{\mathrm{pv}}\hfill \\ \hfill u_{\mathrm{1}}\hfill \\ \hfill u_{\mathrm{2}}\hfill \end{array}\right],

{\mathit{B}}_{\mathrm{1}}=\left\{\begin{array}{ccccccc}\hfill \mathrm{0}\hfill & \hfill -k_{\mathrm{1}}{u}_{\mathrm{1}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill k_{\mathrm{2}}{u}_{\mathrm{1}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill -k_{\mathrm{3}}{u}_{\mathrm{2}}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill -k_{\mathrm{4}}{u}_{\mathrm{2}}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill k_{\mathrm{8}}{u}_{\mathrm{2}}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill & \hfill \mathrm{0}\hfill \end{array}\right\},

{\mathit{B}}_{\mathrm{2}}=\left[\begin{array}{c}\hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill \\ \hfill -k_{\mathrm{5}}\hfill \\ \hfill \mathrm{1}\hfill \\ \hfill \mathrm{0}\hfill \\ \hfill \mathrm{0}\hfill \end{array}\right].

The following transfer functions are derived using Taylors series approximation from Eq. (1) of the state space model to control individual subsystems such as load voltage, battery current, and battery voltage respectively. The transfer function between u₂ and V₀ is expressed as

(2){\displaystyle \frac{V_{\mathrm{0}}}{u_{\mathrm{2}}}}={\displaystyle \frac{k_{\mathrm{8}}{k}_{\mathrm{11}}}{s^{\mathrm{2}}+k_{\mathrm{8}}{k}_{\mathrm{9}}s+k_{\mathrm{8}}{k}_{\mathrm{9}}}},

where

k_{\mathrm{11}}=k_{\mathrm{7}}{x}_{\mathrm{3}}.

The transfer function between u₁ and i_batt is expressed as

(3){\displaystyle \frac{i_{\mathrm{batt}}}{u_{\mathrm{1}}}}={\displaystyle \frac{k_{\mathrm{4}}{k}_{\mathrm{12}}s+k_{\mathrm{3}}{k}_{\mathrm{5}}{k}_{\mathrm{12}}}{s^{\mathrm{3}}+k_{\mathrm{6}}{s}^{\mathrm{2}}+k_{\mathrm{3}}(k_{\mathrm{5}}+k_{\mathrm{2}})s+k_{\mathrm{2}}{k}_{\mathrm{3}}(k_{\mathrm{6}}-k_{\mathrm{4}})}},

where

k_{\mathrm{12}}=k_{\mathrm{2}}{x}_{\mathrm{1}}.

The transfer function between u₁ and V_batt is expressed as

(4){\displaystyle \frac{V_{\mathrm{batt}}}{u_{\mathrm{1}}}}={\displaystyle \frac{(k_{\mathrm{3}}+k_{\mathrm{6}})k_{\mathrm{12}}s-k_{\mathrm{3}}{k}_{\mathrm{4}}{k}_{\mathrm{12}}}{s^{\mathrm{3}}+(k_{\mathrm{6}}-k_{\mathrm{4}}{k}_{\mathrm{2}})s^{\mathrm{2}}+(k_{\mathrm{3}}{k}_{\mathrm{8}}-k_{\mathrm{2}}{k}_{\mathrm{3}}{k}_{\mathrm{5}})s}}.

Problem definition and controller design

Problem definition

The main objective of this paper is to design a powertrain controller that ensure good BMS and extract maximum power from the PV panels to charge the battery. An effective BMS ensures improved battery life and efficiency. Battery parameters such as V_batt, i_batt, and SOC are monitored continuously to maintain these parameters within permissible limits. Hence a HLCA is developed to implement good BMS and extract maximum power. This algorithm switches between various controllers designed to control individual subsystems, such as output load regulation, battery current regulation, battery voltage regulation, and MPPT. BMS is given a higher priority than MPPT during operation of the HLCA. Parametric constraints of the battery, such as i_CT and i_DT, are defined before the execution of HLCA to implement the BMS. MPPT mode is switched on whenever possible by respecting the constraints formulated.

Controller design

Lower level controllers

The complete system is divided into four different subsystems, i.e., output load regulation, battery current regulation, battery voltage regulation, and MPPT of solar PV panels, to ensure good BMS and maximum power extraction.

The speed of the electric vehicle is controlled by the applied voltage of the motor. To achieve this, the transfer function model (2) is used to design a PI controller. The controller parameters are designed by trial and error. Similarly, other objectives are also achieved by considering the transfer functions (3) and (4).

The P&O based MPPT method is used to extract maximum power from the solar PV panels. By sensing V_pv and i_pv of the solar PV array, an algorithm is developed to choose control signal u₁.

High-level control algorithm

The complete structure of the high-level algorithm is represented as a block diagram in Fig. 3. The main objective of the HLCA is to switch between individual lower level controllers based on several criteria to ensure BMS and maximum power extraction. This is explained by the flowchart in Fig. 4. BMS is given a higher priority than MPPT during the course of operation. The battery parameters such as V_batt, i_batt and SOC are monitored continuously to implement the control algorithm.

The HLCA checks all the parameters to switch between controllers. Initially, the MPPT algorithm is switched to generate the control signal u₁ and all the variables are monitored. If the battery starts charging and the limits are respected, the algorithm is continued. By the time the charging current exceeds the limit (|i_batt|>i_CT), the algorithm withdraws MPPT control and switches to constant current charging (CCC) mode. If the same condition occurs with high SOC (SOC>80%), the constant voltage charging (CVC) mode is activated. Similarly, when the battery is supplying additional power, that is, the battery is draining, the MPPT mode is activated. But if the battery drains faster than the limit (i_batt>i_DT), the output regulation is modified to respect the battery constraints.

Simulation analysis

A numerical simulation is performed to evaluate the effectiveness of the proposed controller. The parameters such as SOC, V_batt and i_batt are carefully chosen to execute all the modes of operation. To accomplish the particular operating condition, the load is varied to ascertain the exact parameter values.

The control logic presented in Fig. 4 is used for high-level algorithm implementation. The charging threshold limit of battery current, i_CT is set to 1.25 A. The SOC and load current limits are carefully chosen based on trial and error to execute all the lower level controllers in a sequence for presenting the results. The sequence of operations with its time duration is indicated on the top of Fig. 5 using double side arrow mark. Initially, the system is started with adequate load current to execute the MPPT mode of operation. After some duration, the load is decreased which causes the battery current to increase. Due to this reason, the MPPT is withdrawn, and the CCC mode comes into operation which can be observed in Fig. 5. The MPPT mode is withdrawn due to the violation of battery current i_batt. When it is about to exceed 1.25 A, the mode switching occurs, and the battery continues to charge at the set charging current (CCC mode) which is set to 1.25 A. When the same situation occurs at a high SOC (SOC>80%), the CVC mode is switched which is the third sequence of operation in Fig. 5. When the load is consuming the maximum available solar power and still if the load demand is high, it will start to take power from the battery. In this stage (i_batt>0), the MPPT mode is switched. Furthermore, when the battery is draining more (i_batt>i_DT), the load regulation mode is adapted to respect the BMS. The discharge threshold limit of battery current i_DT is set to 1 A and hence, when i_batt>1 A, the load regulation mode is switched to ensure that the battery is not draining heavily. The constraints formulated for execution of lower level controllers can be found in Fig. 3.

The simulation results are obtained under constant irradiation and temperature conditions. The solar irradiance and temperature are fixed at 1000 W/m² and 25°C while performing the simulation. The developed high-level algorithm switches efficiently to the various lower level controllers based on the instantaneous variation of the parameters. The variation of V_pv, i_pv, V_batt, i_batt, SOC, and i_load are mapped on the same graph for different operating modes.

Experimental investigations

Experimental prerequisites

To execute the proposed control algorithm, a dedicated real-time controller with an integrated development environment (IDE) is chosen for programming. The HLCA is developed to respect the constraints of BMS and maximum power extraction. This is designed using the logic presented in Fig. 4. In addition to several analog inputs/outputs ports, the dedicated controller also possesses differential input port which is helpful in interfacing bidirectional power source such as a battery. The controller is connected to a computer while performing the experimental test. The programming software of this dedicated controller is enriched with graphical user interface (GUI) options which help to control, monitor, and record the real-time values of system parameters in parallel. To implement the control logic, the parameters such as i_batt, V_batt, i_pv and V_pv are sensed using dedicated voltage and current sensors.

The SOC estimation is based on the integral of i_batt considering the capacity of the battery. The formula employed for estimation of SOC is given by

(5)\mathrm{SOC}={\mathrm{SOC}}_{\mathrm{0}}-{\displaystyle \frac{\mathrm{1}}{C}}{\displaystyle {\int }_{\mathrm{0}}^{T}i\left(t\right)\mathrm{d}t}.

where SOC₀ is the initial SOC of the battery and C is the capacity of the battery. Voltage and current sensors used to sense the corresponding voltage and current values are of Hall effect based transducers. These sensors have an excellent accuracy (±0.90%) and good linearity (linearity error<0.2%) characteristics [⁴⁴,⁴⁵]. Sensor circuits are designed based on the operating values of the parameters of the system and are developed in a standard printed circuit board. These sensor boards are calibrated to measure the actual value of the system parameters.

The electrical specification of the individual solar PV module is listed in Table 1, whereas the specification of the solar PV arrays is presented in Table 2. The controller with interfaced sensors is illustrated in Fig. 6. The specifications of lead-acid battery used for the experiment is tabulated in Table 3. To energize a 48 V motor, four numbers of these batteries are used and connected in series. A 250 W, 48 V BLDC motor is used for experimental investigations. A dedicated loading arrangement is developed in the laboratory to flexibly vary the load current of the electric motor to execute all the individual controllers. To accomplish the specific operating condition, the load is changed to ascertain the exact parameter values. In the developed loading arrangement, an electric motor is coupled with a permanent magnet AC (PMAC) generator and is fitted with the same frame. The coupling is made using a belt drive system. Further, the generated AC output of PMAC generator is rectified using a diode bridge rectifier (DBR). The output of DBR is connected across a rheostat which is varied to ascertain the load change. A photographic image of the developed experimental setup along with loading arrangement is depicted in Fig. 7.

Experimental implementation

The details of steps/procedure followed for the experimental execution of proposed powertrain controller is discussed as follows. In the view of executing the entire lower level controller, the load (speed of the electric motor) is varied intentionally at the different operating point to switch between the controllers. The control logic presented in Fig. 4 is followed to develop the controller. Initially, MPPT mode is executed by varying the load current of the system. By increasing the duty cycle u₂ and by varying the rheostat of the loading arrangement, the electric motor is made to run at some critical speed to execute the MPPT mode. During this mode, the constrain i_batt<0 and |i_batt|<i_CT is respected. Now, to switch the control to CCC mode, the load of the system is gradually reduced by varying the rheostat and keeping it to maximum resistance position. If required, the speed of the electric motor can also be brought to lower speed to achieve the constraints of CCC mode, i.e., i_batt<0 and |i_batt|>i_CT . During this mode, the SOC of the battery lies below 80%. In continuation with CCC mode, during the course of operation, when the SOC of the battery exceeds 80%, i.e. (when SOC>80%) the control automatically switches to CVC mode. This mode will be in operation till the constraints i_batt<0 and |i_batt|>i_CT are valid. Now, the load is varied to utilize battery power. This is done by increasing the load of the system near to full load value. During this stage, the battery current becomes positive (i_batt>0) and the controller automatically switches to MPPT mode. When the load is increased further, the battery tends to supply high current |i_batt|>i_DT and the control now switches to load regulation mode.

Experimental results

The experimental results of the parameters of the system during the execution of all lower level controllers are presented in Fig. 8. It can be observed from Fig. 8 that, initially the MPPT mode is executed, the duration of which is indicated by the double-sided arrow on the top of the image. In this mode, the duty cycle is observed to be maximum, which indicates that maximum power is extracted from the solar panels. The parameter SOC is carefully chosen at the beginning of the experiment to execute other modes such as CCC and CVC. The initial value of SOC chosen is 79%. At the end of the first MPPT mode, the SOC level increases and reaches almost 79.5%. The battery charging threshold current i_CT is set to 1.25 A. To switch to the next mode, i.e., CCC, the load of the system is decreased intentionally which causes the battery current to increases its magnitude. When |i_batt| is about to exceed 1.25 A, the control automatically switches to the CCC mode. In the CCC mode, it is observed that the battery charging current is maintained constant at 1.25 A and its voltage magnitude increases slightly. It can be inferred from Fig. 8 that SOC increases in this mode. When SOC is about to exceed 80%, the control automatically switches to the CVC mode. In this mode, V_batt is maintained constant, and i_batt tends to decrease as it takes less current when the SOC level is growing continuously. When the load increases gradually, the battery current i_batt tends to become positive, and the control is now switched to the MPPT mode which is also observed in Fig. 8. The discharge threshold limit of battery current i_DT is set to 1 A. Now, when the load is increased heavily, the battery current tends to drain more, and when it is about to exceed 1 A, the control is switched to the load regulation mode, as indicated in Fig. 8.

Conclusions

A powertrain controller is presented for the solar PV-battery fed HEV. The control strategy is designed to extract maximum power from the solar PV panels whenever possible and manages the power flow in the EV by respecting the constraints formulated in the battery management system. The proposed control algorithm is successfully evaluated by performing real-time experimental studies and a numerical simulation. The high-level algorithm is tested under four different conditions such as low SOC, high SOC, maximum battery current, and heavy load condition. It is noted that the switching between the lower level controllers is achieved corresponding to the operating condition without violating the limits of the battery management system. Hence, the health of the battery and safety are ensured at all operating conditions. Further maximum power extraction from solar photovoltaic panels is also achieved under heavy load condition without violating the limits. The optimal performance of the system is ensured by maintaining good battery health and utilizing maximum power from solar photovoltaic panels. The proposed control algorithm is effective and can be easily adapted to other HEVs by customizing the constraints of the battery management system based on the specifications of the battery. The experimental and simulation results follow a similar pattern for all operating conditions. Both simulation and experimental results are presented for validation.

References

[1] Çağatay Bayindir K, Gözüküçük M A, Teke A. A comprehensive overview of hybrid electric vehicle: powertrain configurations, powertrain control techniques and electronic control units. Energy Conversion and Management, 2011, 52(2): 1305–1313

[2] Fathabadi H. Plug-in hybrid electric vehicles (PHEVs): replacing internal combustion engine with clean and renewable energy based auxiliary power sources. IEEE Transactions on Power Electronics, 2018, 33(11): 9611–9618

[3] Garcia-Valle R, Lopes J A P. Electric Vehicle Integration into Modern Power Networks. New York: Springer Science & Business Media, 2013

[4] Lin X, Lei Y. Coordinated control strategies for SMES-battery hybrid energy storage systems. IEEE Access, 2017, 5(1): 23452–23465

[5] Vinot E, Reinbold V, Trigui R. Global optimized design of an electric variable transmission for HEVS. IEEE Transactions on Vehicular Technology, 2016, 65(8): 6794–6798

[6] Shabbir W, Evangelou S A. Exclusive operation strategy for the supervisory control of series hybrid electric vehicles. IEEE Transactions on Control Systems Technology, 2016, 24(6): 2190–2198

[7] Eckert J J, Silva L C, Costa E S, Santiciolli F M, Dedini F G, Corrêa F C. Electric vehicle drive train optimisation. IET Electrical Systems in Transportation, 2017, 7(1): 32–40

[8] Sun L, Feng K, Chapman C, Zhang N. An adaptive power-split strategy for battery–super capacitor powertrain design, simulation, and experiment. IEEE Transactions on Power Electronics, 2017, 32(12): 9364–9375

[9] Locment F, Sechilariu M, Forgez C. Electric vehicle charging system with PV grid-connected configuration. In: Vehicle Power and Propulsion Conference (VPPC), Illinois, USA, 2011: 1–6

[10] Pourabdollah M, Egardt B, Murgovski N, Grauers A. Convex optimization methods for powertrain sizing of electrified vehicles by using different levels of modeling details. IEEE Transactions on Vehicular Technology, 2018, 67(3): 1881–1893

[11] Naseri F, Farjah E, Ghanbari T. An efficient regenerative braking system based on battery/supercapacitor for electric, hybrid, and plug-in hybrid electric vehicles with BLDC motor. IEEE Transactions on Vehicular Technology, 2017, 66(5): 3724–3738

[12] Xu G, Xu K, Zheng C, Zhang X, Zahid T. Fully electrified regenerative braking control for deep energy recovery and maintaining safety of electric vehicles. IEEE Transactions on Vehicular Technology, 2016, 65(3): 1186–1198

[13] Liu B, Li L, Wang X, Cheng S. Hybrid electric vehicle downshifting strategy based on stochastic dynamic programming during regenerative braking process. IEEE Transactions on Vehicular Technology, 2018, 67(6): 4716–4727

[14] Nian X, Peng F, Zhang H. Regenerative braking system of electric vehicle driven by brushless DC motor. IEEE Transactions on Industrial Electronics, 2014, 61(10): 5798–5808

[15] Fathabadi H. Two novel techniques for increasing energy efficiency of photovoltaic-battery systems. Energy Conversion and Management, 2015, 105: 149–166

[16] Hemmati R, Saboori H. Emergence of hybrid energy storage systems in renewable energy and transport applications—a review. Renewable & Sustainable Energy Reviews, 2016, 65: 11–23

[17] Li X, Hui D, Lai X. Battery energy storage station (BESS)-based smoothing control of photovoltaic (PV) and wind power generation fluctuations. IEEE Transactions on Sustainable Energy, 2013, 4(2): 464–473

[18] Saxena N, Hussain I, Singh B, Vyas A L. Implementation of a grid-integrated PV-battery system for residential and electrical vehicle applications. IEEE Transactions on Industrial Electronics, 2018, 65(8): 6592–6601

[19] Kim H, Parkhideh B, Bongers T D, Gao H. Reconfigurable solar converter: a single-stage power conversion PV-battery system. IEEE Transactions on Power Electronics, 2013, 28(8): 3788–3797

[20] Sivaprasad A, Kumaravel S, Ashok S. Integration of solar PV/battery hybrid system using dual input DC-DC converter. In: 2016 Biennial International Conference on Power and Energy Systems: Towards Sustainable Energy (PESTSE), Bengaluru, India, 2016, 1–5

[21] Shafiei N, Ordonez M, Saket Tokaldani M A, Arefifar S A. PV battery charger using an L3C resonant converter for electric vehicle applications. IEEE Transactions on Transportation Electrification, 2018, 4(1): 108–121

[22] Simoes M G, Franceschetti N, Adamowski J. Drive system control and energy management of a solar powered electric vehicle. In: Proceedings of the 13th Applied Power Electronics Conference and Exposition, Anaheim, CA, USA, 1998, 1: 49–55

[23] Ippolito M, Telaretti E, Zizzo G, Graditi G. A new device for the control and the connection to the grid of combined res-based generators and electric storage systems. In: International Conference on Clean Electrical Power (ICCEP), Alghero, Italy, 2013: 262–267

[24] Rahimi-Eichi H, Ojha U, Baronti F, Chow M Y. Battery management system: an overview of its application in the smart grid and electric vehicles. IEEE Industrial Electronics Magazine, 2013, 7(2): 4–16

[25] Hu K W, Yi P H, Liaw C M. An EV SRM drive powered by battery/supercapacitor with G2V and V2H/V2H capabilities. IEEE Transactions on Industrial Electronics, 2015, 62(8): 4714–4727

[26] Guo X, Li J, Wang X. Impact of grid and load disturbances on electric vehicle battery in G2V/V2G and V2H mode. In: Energy Conversion Congress and Exposition (ECCE), Montreal, Canada, 2015, 5406–5410

[27] Shen J, Khaligh A. Design and real-time controller implementation for a battery ultracapacitor hybrid energy storage system. IEEE Transactions on Industrial Informatics, 2016, 12(5): 1910–1918

[28] Xiong R, Cao J, Yu Q, He H, Sun F. Critical review on the battery state of charge estimation methods for electric vehicles. IEEE Access: Practical Innovations, Open Solutions, 2018, 6: 1832–1843

[29] Graditi G, Adinolfi G, Tina G. Photovoltaic optimizer boost converters: temperature influence and electro-thermal design. Applied Energy, 2014, 115: 140–150

[30] Brahmi H, Dhifaoui R. Dynamic characteristics and improved MPPT control of PV generator. Frontiers in Energy, 2013, 7(3): 342–350

[31] Koad R B, Zobaa A F, El-Shahat A. A novel MPPT algorithm based on particle swarm optimization for photovoltaic systems. IEEE Transactions on Sustainable Energy, 2017, 8(2): 468–476

[32] Peng B R, Ho K C, Liu Y H. A novel and fast MPPT method suitable for both fast changing and partially shaded conditions. IEEE Transactions on Industrial Electronics, 2018, 65(4): 3240–3251

[33] Jeon Y T, Lee H, Kim K A, Park J H. Least power point tracking method for photovoltaic differential power processing systems. IEEE Transactions on Power Electronics, 2017, 32(3): 1941–1951

[34] Piegari L, Rizzo R. Adaptive perturb and observe algorithm for photovoltaic maximum power point tracking. IET Renewable Power Generation, 2010, 4(4): 317–328

[35] Ahmed J, Salam Z. An enhanced adaptive P&O MPPT for fast and efficient tracking under varying environmental conditions. IEEE Transactions on Sustainable Energy, 2018, 9(3): 1487–1496

[36] Mohanty S, Subudhi B, Ray P K. A grey wolf-assisted perturb & observe MPPT algorithm for a PV system. IEEE Transactions on Energy Conversion, 2017, 32(1): 340–347

[37] Mahmoud Y, El-Saadany E F. A novel MPPT technique based on an image of PV modules. IEEE Transactions on Energy Conversion, 2017, 32(1): 213–221

[38] Aurilio G, Balato M, Graditi G, Landi C, Luiso M, Vitelli M. Fast hybrid MPPT technique for photovoltaic applications: numerical and experimental validation. Advances in Power Electronics, 2014, 2014: 1–15

[39] de Brito M A G, Galotto L, Sampaio L P, de Azevedo e Melo G, Canesin C A. Evaluation of the main MPPT techniques for photovoltaic applications. IEEE Transactions on Industrial Electronics, 2013, 60(3): 1156–1167

[40] Subudhi B, Pradhan R. A comparative study on maximum power point tracking techniques for photovoltaic power systems. IEEE Transactions on Sustainable Energy, 2013, 4(1): 89–98

[41] Rezk H, Fathy A, Abdelaziz A Y. A comparison of different global MPPT techniques based on meta-heuristic algorithms for photovoltaic system subjected to partial shading conditions. Renewable & Sustainable Energy Reviews, 2017, 74: 377–386

[42] Kini R L, Sellers A J, Hontz M R, Kabir M R, Khanna R. Comparison of GAN and SI-based photovoltaic power conversion circuits using various maximum power point tracking algorithms. In: Applied Power Electronics Conference and Exposition (APEC), Tampa, USA, 2017, 2977–2982

[43] Aamri F E, Maker H, Sera D, Spataru S, Guerrero J M, Mouhsen A.A direct maximum power point tracking method for single-phase grid connected PV inverters. IEEE Transactions on Power Electronics, 2018, 33(10): 8961–8971

[44] LEM. Current Transducer LA 55-P. 2018–11–20, available at lem website

[45] LEM. Voltage Transducer LV 25-400. 2018–11–20, available at lem website

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature