Optimization of the fuel economy and emissions for plug in hybrid electric recreational boat energy management strategy using genetic algorithm

Received Oct 2, 2018 Revised Oct 19, 2019 Accepted Feb 16, 2019 Today, the transportation sector has undergone a change from conventional vehicle to hybrid electric vehicle especially land-based with the aim to reduce fuel consumption and emissions. However, water transportation is also one of the contributors of excessive use of fuel and emissions. Therefore, water transport needs changes as it has been done on land transport, especially cars. In this paper, plug in hybrid electric recreational boat (PHERB) is introduced. PHERB is a special model because in PHERB powertrain configuration, it only needed one EM compared to existing configuration with energy management strategy (EMS). In this work, the optimal EMS for PHERB are presented via genetic algorithm (GA). To estimate the fuel economy and emissions, the model of PHERB is employed numerically in the MATLAB/SIMULINK environment with a special EMS using Kuala Terengganu (KT) river driving cycle. Simulation result of PHERB optimization using GA improve to 15 % for KT river driving cycles without violating the PHERB performance.


INTRODUCTION
Nowadays, the fuel consumption and emission become big issues in transportation. This issue also effect the water transportation. In water transportation, the boat is used internal combustion engine to drive the boat. Plug-in hybrid electric vehicle give the great promising in reducing fuel economy and emission. Thus, water transportation needs the innovation to overcome these problems [1][2][3][4]. In this paper, Plug-in hybrid electric recreational boat (PHERB) is a new invention of recreational boat was presented [5]. The proposed PHERB is illustrated in Figure 1.
PHERB consists only one electric machine (EM) which functions as either a motor or generator at one time while in the energy system storage (ESS) there is ultracapacitor (UC) bank for fast charging and discharging during the regenerative braking and fast acceleration. The full size of ICE is required because when the state of charge (SOC) of the ESS is low, the ICE will move the boat alone until the ESS SOC reaches a high level and the EM will take over to move the boat. A special energy management strategy (EMS) for the PHERB is needed to improve the fuel economy and emissions. In this paper, the model of PHERB with a special EMS were developed mathematically numerical in MATLAB/SIMULINK is presented. The Kuala Terengganu driving cycle were used in PHERB model to analyse the PHERB performance. Genetic algorithm (GA) optimization were applied to parameter used in PHERB to improve the PHERB model in term of fuel consumption and emission. The minimum of fuel consumption and emission before and after optimize were compared.  [6]. The boat type selected is a recreational boat. In simulation, the length of boat used is 12.4 m and density of water is 1000 kgm -3 . The enlargement of boat model starts with the estimations of boat energy and power needed for common driving situations based on the parameters and target specifications of the boat based on PHERB specification, parameter and requirement. Through a power flow analysis, the size and limit of each component are decided accordingly to get the necessities. Table 2 displayed the size and specifications used for PHERB. All parts in PHERB obtain a mathematical model is combined. The boat performance is simulated in the MATLAB/SIMULINK environment with a special EMS and different driving cycle. Overall structure of PHERB model is illustrated in Figure 2.

PHERB ENERGY MANAGEMENT SYSTEM
Energy Management System (EMS) is dependable of choosing in which mode that the boat is working. A few working methods of the proposed EMS appeared in Figure 3 to control the flow of power among the ESS, EM, and ICE, including the mechanical braking, regenerative braking, motor only, engine recharge, motor and engine, and engine only mode refer to the boat demand in forward and reverse cruising and SOC level of ESS [7][8]. The mechanical braking mode is started if the SOC of both ESS and the throttle position is high. Amid the regenerative braking mode, the distribution of absorbed regenerative power be determined by on the percentage of throttle position as well as on the SOC level of both storage units. EM only mode can be activated when the SOC level is high. When the ESS SOC and the speed are low, the ICE will drive the boat for charging the ESS. If the boat is cruising and the ESS has a moderate SOC, then the boat can be either ICE recharge or EM only mode refer to previous mode. If the boat acceleration is high, then the ICE will not have an opportunity to charge the ESS because the ICE only mode is activated.

PHERB SIMULATION
In simulation, PHERB model presented in Figure 4 is used to simulate the drive performance and fuel consumption for Kuala Terengganu (KT) driving cycle. KT driving cycle are display in Figure 5.   Figure 10 shows the simulated EM speed, torque and power. As shown in the figure, when the boat accelerates, the required EM torque increases quickly, and when the reaches the relatively stable velocity level, a much smaller torque is required to overcome the resistance and water drag to the boat. The average power demand from the EM is 5 kW at the velocity level and the peak power demand is 13 kW during the acceleration boat.   Figure 13 plots the simulated vehicle speed and the demanded speed for the KT drive cycle. As shown, the two curves match very well with a very small error of less than 0.05 m/s between the targeted and acquired speeds. So that, it is expected that the PHERB be able to provide drive performance as the conventional hybrid boat under the actual drive conditions.

PHERB OPTIMIZATION
A process of knowing the minimum and maximum limits of an objective function while at the same time satisfying certain constrains on the design variables and also selecting the best combination in every iteration. In this study, the model in the loop approach are used to design optimization process. Firstly, the PHERB model simulation developed initial values of the design variable, and from this value, the numerical values of the objective function were appear. In this case, the value of fuel consumptions and emission decrease, which is imitated by four parameters including fuel usage, hydro-carbon (HC), carbon monoxides (CO) and nitrogen oxides (NOx) level. The genetic algorithm (GA) is used in the model-based design optimization. It can be summarized briefly as illustrated in the flowchart of Figure 14. Table 3 shows a set of parameters used in this research [9]- [11]. The objective is to minimize the fuel consumption and emissions, such as HC, CO and NOx, of the boat for the KT drive cycle.
The initial value of design variables given in Table 4 are simulated in PHERB model. Table 4 shows the six design variables used in this study. Each design variable is also restricted within a lower and an upper bound were stated in Table 5. The fuel consumption and emissions were observed for KT drive cycle. A comparison of the fuel economy and emissions before and after the optimization is given in Table 6. A significant improvement in the fuel used and emissions is achieved by optimization.    Table 7 shows the final values of the six design variables after optimization. It is shown that the rating of the EM is greatly reduced, implying that down-sizing of the EM has been achieved. On the other hand, the ICE is down-sized to a lesser extent.

CONCLUSION
The results of boat subsystems in terms of ESS current, voltage and output power, and EM speed, torque and power, are within reasonable and expected range. The components of the boat subsystems are correctly sized to achieve the good performance at target velocity. It can be concluded that results of the PPHERB model are correct. Based on the optimization results, the following observations can be made. The fuel used of the PHERB is decreased and the emissions of HC, CO and NOx are decreased as well with the GA optimization. The power rating of the ICE and EM are reduced significantly.