5.1.4.Simulation of supercapacitor energy storage system
The efficiency of a hybrid ESS with supercapacitor and accumulator batteries strongly depends on the choice of the control method of energy flow through the storage batteries and on the possibility to absorb the regenerative energy from other vehicles. For prolonging of traction accumulator battery lifetime no peak power shaving from accumulator is allowed, traction battery is manually switched to traction mode if overhead network voltage is unavailable. The traction battery charging with low current should be maintained for a long time independently of the vehicle operation mode. It is advisable to provide continuous charging of the accumulator battery even at frequent overhead voltage dropouts caused by a light rail vehicle running through the overhead section insulators and crossings with trolleybus lines. However, electrically operated rail and overhead switches may require the vehicle to cross them without drawing power from overhead line interrupting charge process. Therefore, rather poor peak shaving and charging requirements make the ESS structure with an conventional accumulator only and without a supercapacitor less attractive.
The ESS control strategies are depending of the goal of ESS using, which could be saving of regerative braking energy, reducing of traction motor starting peak power, network voltage stabilization and autonomous running without connection to external power source. The common ESS control methods [14] are mean power control strategy, mean power control strategy with prediction, power supply voltage stabilization, peak power shaving and proportional control.
Mean power control strategy is control method when consumer tries to draw constant power from the overhead line covering the fluctuating power part with the help of ESS. In the upgraded variant of this method - mean power control strategy with prediction, constant power is calculated and maintained for small period of time, few minutes instead of day and by taking in account time and vehicle displacement for prediction of consumer power. Theoretical minimum in the energy consumption is achieved for an ideal ESS with a sufficiently large energy capacity which could be too overdimensioned for on-board ESS. Due to limited capacity and lifetime of accumulator battery and necessity of permanent connection to power network, both mean power control methods are not very suitable for vehicle mobile ESS applications especially in the case of autonomous traction. The voltage stabilization strategy provides ESS operation as additional source and load depending from converter input DC link voltage. The voltage stabilization method is acceptable for small power generators and ESS in autonomous feeding mode to provide the constant voltage source feeding of load. The main idea of the peak shaving control strategy is to store braking energy and to discharge ESS only when the tram power is higher than the certain level. The main idea of the proportional control strategy is to store braking energy and to discharge ESS with the power proportional to that of tram traction. The substation current could be decreased with retaining the ESS energy for longer drive running. The energy distribute factor between substation and supercapacitor is 50% of ESS current. Variant of such method described in [8,15] suggests to start the traction motor from ESS only (100% supercapacitor ESS current) in both network and autonomous supply modes, but drawback of this variant is steep increasing of power substation current at the end of ESS discharge, when supercapacitor becames empty, therefore such control variant does not provide enough power for autonomous traction with supercapacitor ESS only, hybrid ESS is required.
To check whether a vehicle with ESS is able to safely use regenerative braking energy, its operation at various voltage failures and short dropouts in the network should be studied [1,7,12]. It is also necessary to compare how the vehicle drive and its ESS system operate at accumulator continuous charging and with switched off accumulator. As the accumulator is not operated in parallel connection to network power supply, the accumulator discharge mode is not observed.
For undisturbed regular traffic, experiments could be run only on a limited scale, using special test tracks or regular routes in the off-peak time. Although at the final stage of manufacturing and certification the obligatory tests of a vehicle are conducted, the number of test runs could be reduced by adjusting the drive TB and computer models instead of full-scale testing of vehicle equipment. To test the traction drive system a TB for the electric drive system with a hybrid ESS is proposed. However, even in a low power range the complexity and price of hybrid ESS equipment remain relatively high, which requires additional protection from possible abnormal operation as well as careful pre-operation tests for fault prevention to be provided. Therefore, it is advisable to investigate the operation of scale TB instead of vehicle, but proper design of bench requires first to build and test a computer model of the proposed TB, which would allow its safe testing and upgrading before physical implementation [16,17].
When studying an ESS we should compare the operation of light rail vehicles (e.g. trams) equipped and not equipped with such an energy storage system in different operation and supply modes: drive and braking, with the network and autonomous power supply. The drive is meant to continue normal running during the automatic switching between these supply modes at any operation mode of the traction motor, including the most difficult case when the accumulator has been charged at the highest acceleration rate of a full-loaded vehicle.
The operation time scale of our stationary TB is 1:1 with respect to
a real tram factory test diagram. As the energy capacity
of a typical ESS is limited by technical and economic reasons, the
maximum speed of a vehicle in the autonomous traction mode is as low
as 30...40 km/h (with full traction DC motor excitation) to prevent
fast full discharge of ESS batteries. The field weakening of a
traction motor should not be used in any battery supply mode;
therefore, correct comparison of a conventional tram with an
ESS-equipped one could be done at the normal performance of the
traction motor [12]. For this reason, the TB speed scale is chosen
ignoring the field reduction factor, with the nominal speed of 1370
RPM assumed as that of the tram traction motor. The TB studies were
carried out at the fixed maximum traction current of 40A, which
corresponds to maximum traction current of the tram.
A simplified diagram of the traction drive TB with a hybrid ESS is shown in Fig. 6. The vehicle drive is simulated by a traction drive model containing an equivalent DC motor with independent excitation. The DC motor is connected to the 110V DC bus through a two-quadrant pulse width modulated (PWM) DC/DC converter, which allows simulation of the traction and regeneration braking modes without reversing the rotational direction. The DC bus is connected to the 110V DC power supply through a diode VD simulating a unidirectional traction substation, with no regeneration energy return possible. The DC bus is equipped with a large capacitor Cf − equivalent to the filter capacitor of vehicle traction
Fig. 6. Simplified diagram of the traction drive test bench with a hybrid energy storage system
drive. The hybrid ESS contains supercapacitor and accumulator batteries connected to the DC bus through bidirectional DC/DC converters.
In the braking mode of the traction drive the total braking energy should be transferred to the hybrid ESS. In the cases when it is impossible (e.g. the ESS is fully charged), a DC bus overvoltage can occur. To avoid this, the voltage limiter is connected to the DC bus. It limits bus voltage to the set value 125V dissipating excess energy in the rheostat Rvl.
The DC motor, which simulates the electric transport traction drive, is coupled to an AC induction motor driven by a frequency converter FC; this, in turn, simulates the traction drive load. The FC is connected to the 380V/50Hz AC network and has no regenerative braking controllable rectifier. The AC drive operates in the braking mode as the load for the traction drive DC motor operating in the drive mode. The braking energy produced by the load simulator is transferred (by a braking chopper embedded in the FC) to the braking resistor Rbr chosen for long-term continuous operation.
The control system controls the FC in a way that the AC motor acts as a rotating object with a very high moment of inertia simulating a large-mass vehicle.
The control system of a supercapacitor ESS converter in the charging mode it acts as a DC bus voltage stabilizer at a set value of 120V. In the discharging mode it normally draws current according to the reference value set by the central controller. In case of input voltage dropout it maintains the DC bus voltage at the 95V level. In both the modes of
Fig. 7. Simplified Matlab/Simulink model of test bench
operation the limitation of supercapacitor current is activated with a set value ±45A. The system for voltage monitoring of the supercapacitor protects it against overvoltage.
Controller of the battery converter is simpler than supercapacitor converter controller. Two constant charge current values, 1.8A and 4.8A, for slow and fast battery charging can be set. Discharge current is set by the central control unit. By monitoring the battery voltage (state) the local controller enables or disables the charging or discharging modes.
A simplified Matlab/Simulink model of the TB is shown in Fig. 7. The controlled voltage source CVS and diode D1 simulate the feeding substation, while R1 is an equivalent resistance of the overhead line.
The DC bus voltage limiter is composed of DC voltage source V2 and diode D3. Two signal builder blocks I_ref and V_sub are used as reference signals for the traction current and substation voltage, respectively, to simulate different operation modes (running/braking, substation voltage dropouts).
The traction DC motor is coupled to a load simulator. Since only the battery charge mode is simulated, its converter is replaced with a controllable current source I_bat.