The remainder of this paper is organized as follows: Section 2 presents an overview of the powertrain structures and characteristics of EWLs. Section 3 introduces the experiment measurement for a dual-motor distributed drive EWL. Section 4 provides the experimental results, and Section 5 discusses the experimental results. Section 6 concludes the paper.
The literature discussed above highlights the advantages of using multi-motor drive methods to achieve an optimal power distribution and flexible control strategies, leading to improved dynamics and economy in electric drive vehicles. Similar considerations apply to EWLs, which also need to operate efficiently and dynamically. As a result, improving the drive structure and control strategies of EWLs has emerged as a key research topic in this field. EWL drive systems can be broadly categorized based on the number of motors, namely a single-motor drive system and a multi-motor drive system. A multi-motor drive system can further be classified as either a multi-motor independent drive system or a multi-motor coupled drive system [ 31 ]. This study takes the multi-motor independent drive system of an EWL as the research object to compare its energy conversion efficiencies in three drive modes.
Additionally, Xu, S. et al. [ 29 ] put forward a multi-mode drive optimization control strategy based on a hierarchical control architecture to investigate the economy and dynamics of an electric vehicle with dual motors in the front axle and a single motor in the rear axle. The test showed a 6.45% reduction in energy consumption. A two-in-one motor drive control strategy is proposed that takes into account air-conditioning usage conditions based on factors such as motor efficiency within the two motors [ 30 ], which do not have the same maximum power, and its simulation results show that energy savings of up to 2.2% can be achieved.
The drive structure of an EWL can be referred to as the structure of an electric road vehicle [ 13 ]. As shown in Figure 1 , a pure electric drive vehicle structure is mainly powered by the power battery, and it is driven by the electric motor through a variable speed gearbox, differential drive axles, and half shafts to realize the transmission of torque and speed to the drive wheels. As far as the arrangement of motor drives is concerned, the types that are being researched include single-motor drive [ 14 ], simultaneous drive with dual motors [ 15 ], dual-motor drive with all 4 wheels [ 16 ], hub motor drive [ 17 ], etc. In terms of applications, single-motor drives with fixed reduction ratios are the most widely used in electric vehicles due to the simplicity and ease of control of the single-motor drive system. This reduces the production costs of electric vehicles and facilitates maintenance [ 18 ]. In order to improve the efficiency and performance of electric drives, researchers have made improvements to the drive structure of single-motor drives, such as utilizing automated manual transmission (AMT) [ 19 20 ], two-speed automatic transmission (AT) [ 21 22 ], continuously variable transmission (CVT), and dual-clutch transmission (DCT) [ 23 25 ]. However, the use of a single motor does not allow the vehicle to operate in the economic speed zone of the motor for long periods of time, which is detrimental to increasing the range of the vehicle. Therefore, research into multi-motor drive structures is also considered necessary to address this issue. Wang, Y. et al. [ 26 ] presented a new two-motor hybrid drive system with two power sources, as shown in Figure 2 . The system achieves torque–speed coupling between the two power sources, greatly increasing the high-performance operating range of the motors. At the same time, a CVT is implemented to effectively increase the driving range. Holdstock, T. et al. [ 27 ] designed a two-motor, four-speed electric drive architecture, as shown in Figure 3 . This drivetrain architecture improves the flexibility of the motor operating points. The results of the study show that under the ECE-15 operating conditions test, the drive efficiency of the single-row planetary gear structure increased by 9.1%, and the energy recovery efficiency increased by 9.7%. Additionally, the drive efficiency of the double planetary gear structure increased by 10.9%, and the energy recovery efficiency increased by 11.1%. Mantriota, G. et al. [ 28 ] proposed a scheme to drive the vehicle with two smaller motors through a planetary gear mechanism, as shown in Figure 4 . The simulation results show that this dual-motor power transmission is usually better than that of a single motor, with an improvement of about 9% in both the drive efficiency and energy recovery efficiency.
A wheel loader (WL) is an off-road vehicle driven by an engine or electric motors and is widely used in specific locations with a small operating area such as construction sites, coal mines, and ports. With the end of the COVID-19 epidemic, the world economy is recovering rapidly. From a developmental perspective, the demand for WLs in various countries has gradually increased. According to a research report [ 1 ], the size and share of the global WL market is predicted to grow to around USD 19,568.7 million by the end of the next five years. Currently, engine-driven loaders still dominate the market, but the environmental pollution caused by economic production activities cannot be underestimated [ 2 3 ]. With the increased severity in environmental protection requirements [ 4 ], the trend towards lower cost and higher efficiency batteries [ 5 7 ] and the development of motor drive control [ 8 10 ] electric wheel loaders (EWLs) will certainly become the mainstream of the future market [ 11 ]. EWLs can reduce noise and exhaust emissions, and the low-speed, high-torque mechanical characteristics of the motor are better suited to the low speeds at which WLs often operate [ 12 ].
When the loader is working at the first stage [ 32 ], the operator uses the maneuvering mechanism to drop the bucket to the ground in preparation for shoveling. However, the operator may not be able to accurately sense the position of the bucket, which can cause the bucket to drop too much, leading to the lifting of the front wheels. This, in turn, can result in insufficient pressure on the ground and reduced traction. Taking into account this driving characteristic of the loader, this study utilizes the EWL configuration shown in Figure 7 for testing. This study conducts bulldozing tests in single front drive, single rear drive, and dual drive modes to investigate which drive mode is the most economical.
In order to solve the problem of the insufficient driving power of a single motor, dual-motor drives are widely utilized in wheel loaders [ 35 36 ]. A type of dual-motor drive structure of an EWL is shown in Figure 6 . The structure of the dual-motor drive for the loader involves the front and rear motors transmitting power to their respective drive axles via a reduction mechanism. Both motors are arranged longitudinally, and their output shafts are connected with a coupling, ensuring that the actual output speed of both motors remains the same at all times when driving. This allows for a power transfer between motors when one is underpowered, resulting in a combined drive. The disadvantage of this construction, however, is that the loader is prone to generating parasitic power. The relevant parameters for each motor and the corresponding drive wheel should satisfy Equation (2) in order to avoid the generation of parasitic power.whererepresents the angular speed of the front motor rotor,represents the angular speed of the front motor rotor,indicates the radius of the front wheel of the EWL,indicates the radius of the rear wheel of the EWL,is the total transmission ratio from the front motor to the front wheel drive train, andis the total transmission ratio from the front motor to the rear wheel drive train.
The advantage of this structure is that there are fewer drive motors, and the control of the motors is simpler. However, the disadvantage is that the front and rear torques are unevenly distributed, and the simple structure of the transfer case cannot distribute the torque according to the front and rear axle torque requirements. As a result, complex transfer case structures and control algorithms need to be developed to compensate for these deficiencies [ 34 ]. Furthermore, due to the high driving forces required for the WL’s running and shoveling conditions, motors that are capable of providing sufficiently high torque and power must be installed.
The working process of a WL usually consists of a four-stage cycle consisting of shoveling, delivery, unloading, and returning to the stockpile [ 32 ]. For the EWL, the drive motor drives the vehicle during both stages of delivery and returning to the pile, which can be referred to as the running condition, where the drive motors contribute the main energy consumption of the WL. Tests have been conducted on the energy consumption of a dual-motor-driven loader in the running condition [ 33 ]. The results indicate that during the running condition, the rolling resistance moment that needs to be overcome by the drive motor is the same, whether it is driven by the front wheels alone or by the rear wheels alone, which depends on the total mass of the WL and the rolling resistance coefficient of the road surface. In the running condition, the EWL can be driven by either a single motor or two motors. Figure 5 shows a single-motor drive powertrain, which has a similar transmission system to an engine drive except for the gearbox. In this case, a conventional diesel-driven loader is retrofitted with an electric motor in place of an engine. The motor’s speed is reduced by a reduction gear to increase the torque, which is then transferred to the front and rear axles, respectively, via a splitter box. In this configuration, the torquegenerated by the front axle and the torquegenerated by the rear axle and the motor torquehave to satisfy Equation (1).where, andrepresent the angular velocity of the front axis, rear axis, and motor, respectively.is the transmission efficiency of the gearbox.
In practical testing, the measured values of the current, voltage, torque, and speed are not directly represented by functions, and the data collected are discrete points. Therefore, the calculation of the input work cannot be integrated and can only be approximated using Equation (7).
When calculating the collected motor data during testing, the following method should be used: when the loader is moving forward, if the front motor is in the driving state, the torque and current of the front motor are both positive; if the rear motor is in the driving state, the torque of the rear motor is negative, and the current is positive. When the loader is reversing, if the front motor is in the driving state, the torque of the front motor is negative, and the current is positive; if the rear motor is in the driving state, the torque of the rear motor is positive, and the current is positive. In any case, when the current of the motor is positive, it means the motor is working externally, while a negative current indicates that the motor is generating power by being dragged.
In order to ensure the stability of the test data, an experiment was designed for the loader to travel on a horizontal concrete road, and a motor speed of 600 rpm was selected as the target speed. The attitude of the bucket (height and bucket angle) was changed to simulate a realistic scenario of the loader inserting itself into the stockpile before digging the material. Specifically, in the test of the loader’s bulldozing, it was driven forward and backward three times in each of the five states where the front wheels were slightly on the ground, slightly lifted, lifted more, slightly off the ground, and lifted more off the ground. The motor was allowed to operate in three modes: independent front drive, which can be called F-drive; independent rear drive, which can be called R-drive; and dual drive, which can be called D-drive. To facilitate identification, L1, L2, L3, L4, and L5 were used to represent the five position states of the front wheels of the loader, as shown in Table 3
A work efficiency MAP of the motor can be obtained by the data in Table 2 . In the following research, the map is important for the development and optimization of the control algorithm. We draw from the working efficiency MAP that the high efficiency work range is broad.
On this EWL, two PMSMs are utilized as the drive motors. The efficiency of this type of motor under different torques at speeds of 500 rpm, 1000 rpm, 1500 rpm, 2000 rpm, 2500 rmp, and 3000 rpm, with a working voltage of 540 V, are tested and recorded in Table 2
During the test, the control programs for the EWL’s driving and shoveling functions are compiled using Matlab/Simulink. The VCU controls the drive motors, which receive a frequency and current output corresponding to the accelerator pedal and gear signals, thereby driving the EWL. The VCU’s application layer software is also developed using Matlab/Simulink. This software can set three different drive modes for the front motor, rear motor, and dual motor, as well as adjust the target speed of the front and rear motors.
The testing system hardware comprises several components, including the Vehicle Control Unit (VCU), Front Motor Control Unit, Rear Motor Control Unit, Hydraulic Motor Control Unit, Battery Management System, Upper Computer, Data Storage Unit (DTU), and Heat Management Unit. The Upper Computer collects information on the vehicle’s status and sends control information to the VCU to manage the vehicle’s motion. The DTU functions as a storage device that gathers data via the CAN bus. All hardware components communicate via the CAN bus, which has a baud rate of 250 kb/s and uses two terminal resistances of 120 Ohms. The network topology of the hardware system is depicted in Figure 9
A type of distributed electric wheel loader (DEWL) with a battery voltage of 540 V, as shown in Figure 8 , is selected as the test subject in this research. The DEWL is equipped with two electric motors for driving and one electric motor for the work of hydraulic pump. The two driving motors have the same size and parameters of rated power and torque. Numerous parameters influence the energy consumption of electric vehicles during travel [ 38 ]. The specific parameters of the EWL are shown in Table 1
The comparison of Figure 14 Figure 15 and Figure 16 indicates that the motor power fluctuates most sharply in the F-drive mode during the motor acceleration phase, and the fluctuation is more prominent in the rear motor than in the front motor. These factors contribute to the excessive energy consumption in the electric loader operation and need to be avoided.
In L4 and L5, the motor power continues to gradually increase and ranges between 20 kW and 40 kW. Figure 16 illustrates a significant fluctuation in the output torque of the motor when both motors are driven simultaneously. At L1, the power of the front motor is negative, and it is consistent with the torque curve, indicating that the rear motor back-tows the front motor to generate power under this operating condition. From L2 to L4, the power of the front motor tends to decrease gradually, while the power of the rear motor increases gradually.
As can be seen in Figure 14 , the output power of the front motor is increasing as the front wheels are gradually lifted, and the tendency for the rear motor output to increase with the degree of lifting of the front wheels is even more evident. The similar rule of the motor output power can also be observed in Figure 15 . In the L1 condition, the motor power in the smooth phase belonging to the target speed range is essentially less than 20 kW, while in the L2 condition, this power is essentially close to 20 kW, and in the L3 condition, the motor power is larger than 20 kW.
The curves of the input powerand the output powerare plotted in the same graph when the EWL is driven by one motor alone. Figure 14 shows the motor power curves when the EWL is driven by the front motor alone, and Figure 15 shows the motor curves when the EWL is driven by the rear motor alone. When the loader is driven by two motors at the same time, only the output powerof both motors is plotted in the same graph, as shown in Figure 16
The data collected from the tests conducted in the five states from L1 to L5 were collated, and the curves can be classified by the torque–speed and power curves. The tests of F-drive, R-drive, and D-drive can be fully realized in the test conditions of L1, L2, and L3. In the L4 condition, the wheel loader cannot be tested with the F-drive mode, but some of the available data can still be obtained with the D-drive mode. In the L5 state, only the D-drive mode tests can be carried out. The F-drive mode cannot be carried out in L4 and L5 because the ground does not provide sufficient adhesion. In the curves shown in Figure 9 Figure 14 and Figure 15 , FMCU_Sped indicates the front motor speed in rpm; RMCU_Sped indicates the rear motor speed in rpm; FMCU_Tor indicates the front motor torque in Nm; and RMCU_Tor indicates the rear motor torque in Nm. The power curves plotted were calculated from the measured motor current, voltage, speed, and other parameters. RMCU_Prepresents the output power of the rear motor conducting external work, and the instantaneous output power is calculated using Equation (9); RMCU_Prepresents the input power of the motor after the current drive, and the instantaneous input power is calculated using Equation (10).
η P o _ M o t = P ¯ 1 _ P o P ¯ 2 _ P o × 100 %
(11)
P o in the subscript indicates the position of either the front motor by F or the rear motor by R.Due to the target speed control mode being selected in the test, the acceleration and deceleration phrases are not considered in this research. The data for the analysis are taken from Figure 11 Figure 15 and Figure 16 when the motors reach the target speed. The torque of the motor, the output powerof the motor, and the input powerof the motor in the data segment, showing that the torques’ variations are relatively smooth, were selected for the analysis. In the following tables,represents the mean value of torque generated by the rear motor,represents the mean value of output power of the rear motor, andrepresents the mean value of input power of the rear motor, whilerepresents the energy conversion efficiency of the rear motor in the corresponding data segment. Similarly,, andrefer to the average torque, the output power, the input power, and the energy conversion efficiency of the front motor, respectively.indicates the sum average torque of the front and the rear motors. The energy conversion efficiency of the motor can be calculated using Formula (11).wherein the subscript indicates the position of either the front motor by F or the rear motor by R.
The related data calculated in different cases when the EWL was driven by the front motor are listed in Table 4 . The maximum motor efficiency is 52.3% in a forward state and 54.39% in a backward state. Though the F-mode operations have not been realized in the L4 and L5 cases, the data in the three groups of L1, L2, and L3 from Table 4 still illustrate the same pattern of change, as drawn from Table 5
The related data calculated in different cases when the EWL was driven by the rear motor are listed in Table 5 . They show that the energy conversion efficiency of the motor reaches up to 85.43% during the first forward-moving period in the L1 case and the top efficiency is 86.02% during the forward-moving periods in cases L1 to L5. The minimum energy conversion efficiency of the motor is 51.87% in the forward-moving periods, while the lowest is 52.74% in the backward-moving periods. Of all the values of the average torque, the positive values occur in the forward operations and the negative values occur in the backward operations. From L1 to L5, the absolute values of the average torque, average input power, and average output power increase consistently in general, while the opposite is true for the energy conversion efficiency of the motors.
η D _ m o t = ∑ j k T F _ i · n F _ i + ∑ j k T R _ i · n R _ i ∑ j k U F _ i · I F _ i + ∑ j k U R _ i · I R _ i × 9.55 × 100 %
(12)
The related data calculated in different cases when the EWL was driven by the two motors are listed in Table 6 . The necessary parameters of the D-drive mode are the average torque of the motors, the overall energy conversion efficiency of the motors, and the sum torque of the two motors. In the data processing, the sum of the torques of the motors are converted into positive values for the depiction of the curves. The total efficiency of the twin motors is calculated using Equation (12).
Figure 17 displays the average values of the motor torque, input power, output power, and efficiency in F-drive mode from condition L1 to condition L3. Figure 18 presents the average values of the motor torque, input power, output power, and efficiency in D-drive mode from condition L1 to condition L5.The orange curves represent the change in the motor output power, the blue curves show the change in the motor input power, the red curves show the change in the average torque, and the green curves show the change in the motor efficiency. The diagram includes eight curves in both Figure 17 and Figure 18 , in which the solid lines indicate the curves for each parameter when the EWL is moving forward, while the dashed lines indicate the curves for each parameter when the EWL is moving backward.
As shown in Figure 17 , the average torque, average input power, and average output power of the motor in the forward state of the EWL are the smallest in the L1 condition and the largest in the L3 condition. It can be drawn by the change in torque that the resistance rises as the vertical force between the bucket and the ground increases. The energy conversion efficiency of the motor is 50.80% in the L1 condition, 43.75% in the L2 condition, and 40.39% in the L3 condition, indicating that the energy conversion efficiency of the motor decreases as the front wheels of the loader are lifted more. This pattern is consistent for both the forward and backward movements of the EWL.
Figure 17. Curves of motor power, motor efficiency, and motor torque in F-drive mode.
Figure 17. Curves of motor power, motor efficiency, and motor torque in F-drive mode.
As shown in Figure 18 , the average torque, average input power, and average output power of the motor in the forward state of the EWL are the smallest in the L1 condition and the largest in the L5 condition, and all show a gradual increasing trend. The energy conversion efficiency of the motor is 82.62% in the L1 condition, 64.96% in the L2 condition, 58.40% in the L3 condition, 56.03% in the L4 condition, and 52.99% in the L5 condition, indicating a same pattern of change as that in the F-drive mode. The difference is that the motor efficiency is higher in all operating conditions corresponding to the R-drive mode compared to the F-drive mode.
Figure 18. Curves of motor power, motor efficiency, and motor torque in R-drive mode.
Figure 18. Curves of motor power, motor efficiency, and motor torque in R-drive mode.
Figure 19 shows the curve of the variation in the average value of the total motor torque and overall motor energy conversion efficiency from the L1 to L4 operating conditions in the D-drive mode. Shown in purple is the curve of the change in the sum of the torque of the two motors, and shown in green is the curve of the change in the overall motor efficiency.Figure 19 indicates that in the D-drive mode, the sum of the motor torques tends to rise with the increasing bucket ground pressure, whether the EWL is driving forward or backward. When driving forward, the total torque is 174.59 Nm at the L1 condition and 577.7 Nm at the L4 condition, while the energy conversion efficiency is 40.05% at the L1 condition and 82.67% at the L4 condition, which also show an increasing trend. The same pattern of change is observed in reverse. Compared to the F-drive and D-drive modes, this shows that the overall efficiency achieved with the dual-motor drive is higher when the EWL requires more torque.
Figure 19. Curves of total motor torque in D-drive mode.
Figure 19. Curves of total motor torque in D-drive mode.
Based on the analysis above, we can derive a trend table illustrating the changes in the motor torque and motor energy usage efficiency as the height of the front wheel lifting increases, as shown in Table 7
Based on the discussion above, a proposal for the motor energy improvement of dual-motor drive EWLs is put forward below, which may provide researchers and manufactures with a potential solution.
For a two-motor drive EWL, the front wheel can be driven by a smaller motor, while the rear wheel can be driven by a lager one in torque and power. In the running state of shoveling work, the rear motor can be the primary drive motor. When the bucket is lowered towards the ground for the preparation of the shoveling material, this will cause an increase in the pressure on the rear wheel. If the drive force of the rear motor is not enough, the front motor will be activated by the MCU to generate enough toque together with the rear motor. However, this should have a limited condition, that is, the rear motor should work on the high efficiency range as much as possible according to the energy efficiency MAPs of the two motors derived by the tests. In addition, elevating the front wheels to a certain extent can induce tire slippage. When the grip between the front wheels and the ground diminishes, a portion of the energy from the driving force is diverted into parasitic power, an undesirable occurrence. This can be identified by monitoring whether the front wheel rotation speed suddenly exceeds that of the rear wheels. Detecting the tire slip trigger control of the hydraulic circuit governing the bucket or boom allows for adjustments to be made to the bucket’s position and, consequently, allows for traction to be restored between the front wheels and the ground. Employing these methods can effectively reduce energy wastage during loader operations. It is worth noting that various other factors, including the road conditions, terrain, temperature, weather, material properties, and driver behavior, also play roles. These aspects can be considered in future research endeavors aimed at achieving intelligent assisted driving for loaders.
With high quality products and considerate service, we will work together with you to enhance your business and improve the efficiency. Please don't hesitate to contact us to get more details of Electric Heavy Duty Wheel Loader.