GB2500944A

GB2500944A - Series hybrid vehicle

Info

Publication number: GB2500944A
Application number: GB1210771.0A
Authority: GB
Inventors: Adam Chapman; Leon Rosario
Original assignee: Lotus Cars Ltd
Current assignee: Lotus Cars Ltd
Priority date: 2012-02-27
Filing date: 2012-06-18
Publication date: 2013-10-09
Anticipated expiration: 2032-06-18
Also published as: GB201203357D0; GB201210771D0; GB2500944B

Abstract

The invention relates to a series hybrid vehicle, such as an automobile. The vehicle comprises an energy management system, a battery 200, an auxiliary power unit 100 (APU), a load 300 comprising an electric motor 310, and an electrical power bus 400. The vehicle has a first operating mode in which the energy management system determines a target power demand, calculates an APU cost function representing the cost of supplying power from the auxiliary power unit, calculates a battery cost function representing the cost of supplying power from the battery, and calculates a desired APU power to be supplied to the load by the auxiliary power unit and a desired battery power to be supplied to the load by the battery so as to minimise the weighted sum of the APU cost function and the battery cost function to improve fuel efficiency. The invention is aimed at minimizing costs per unit of energy delivered, such as monetary cost, fuel amount, undesirable emissions, component fatigue and/or battery life reduction.

Description

SERIES HYBRID VEHICLE

The present invention relates to a series hybrid vehicle, such as an automobile, and a method of operating the same.

In a series hybrid automobile, the wheels are driven only by electric motors. Power is delivered to the electric motors from a battery and/or from a generator driven by an internal combustion engine.

Internal combustion engines have an efficiency map that varies in dependence upon engine speed and the torque provided by the engine. Therefore, there is a locus of operating points along which the engine can be run to maximise its fuel efficiency per unit of output power.

A simple on/off control strategy has been suggested in which the engine only runs at the optimum operating point. When the power demand for the vehicle is lower than that provided by the engine, the engine can drive the vehicle and also charge the battery. Alternatively, if the battery is sufficiently charged, the engine is deactivated and the battery powers the vehicle. When the power demand for the vehicle is higher than that provided by the engine, the engine can be supplemented by the battery.

Such a control system focuses only on the instantaneous efficient operation of the engine. It can be shown that optimum efficiency of a series hybrid vehicle as a whole can be improved by use of a different control strategy that also considers the cost of charging and discharging

I

the battery over the entire journey of the series hybrid vehicle.

In practice, the whole journey of the vehicle is not known in advance and therefore there is a need to provide a control strategy for a series hybrid vehicle that provides efficient use of fuel in real-time. Moreover, it is necessary to consider the requirement to sustain a state of charge in the battery.

According to a first aspect of the invention, there is provided a method of operating a series hybrid vehicle.

According to a second aspect of the invention, there is provided a series hybrid vehicle.

The term average, as used herein, should be taken to be the mean.

The present invention will now be described, byway of example only, with reference to the accompanying drawings, in which: Figure 1 shows a schematic representation of a series hybrid electric vehicle; Figure 2 shows a basic circuit representing the hybrid electric vehicle; Figure 3 shows how power demand is calculated using current and voltage measurements; Figure 4 shows a graph indicating the efficiency of an internal combustion engine as a function of engine torque and engine speed; and Figure 5 shows a flow chart of the method carried out by the energy management systems.

Figure 1 shows a schematic representation of a four-wheeled automobile embodying a series hybrid electric vehicle of the present invention.

The automobile of Figure 1 comprises an auxiliary power unit 100; a battery 200; a load 300; and a power distribution bus 400.

The auxiliary power unit 100 comprises an internal combustion engine 150 and a generator 110. The internal combustion engine 150 receives a supply of fuel 155 from a storage tank (not shown) Optionally, the generator 110 can act as a starter motor for the internal combustion engine.

The internal combustion engine 150 is arranged to drive the generator 110 which thereby generates electricity and thereby provides electrical power to the power distribution bus 400.

The battery 200 can be temporarily connected to an electrical power supply grid 215 external to the vehicle, via a charger 210 (which may form part of the vehIcle or be external thereto) . In this way the level of charge on the battery may be increased.

The battery 200 is arranged to supply power to the power distribution bus 400 to power the load 300, and is further arranged to receive power from the power distribution bus 400.

The load 300 comprises electric motors 310 arranged to propel the vehicle and, optionally, auxiliary loads such as any heating systems, air conditioning systems, or other electricity dependent systems forming part of the vehiole. In the example shown in Figure 1, the load 300 comprises a pair of electric motors 3101, 31Cr connected one each to the left and right rear wheels 3401, 34Cr of the four-wheeled automobile via a respective gearbox 3301, 33r.

Preferably, the motors 3101, 31Cr can act also as generators 3101, 31Cr. Such generators 3101, 31Cr can be driven by the wheels 3401, 340r under braking and thereby generate electrical power to be provided to the power distribution bus 400 for charging the battery 200.

The power distribution bus 400 is a DC bus. The generator 120 may be an AC generator, in which case an inverter 120 is provided between the generator 110 and the power distribution bus 400. Similarly, the motors 3101, 31Cr may be AC motors, in which case inverters 3201, 32Cr are provided between the motors 3101, 31Cr and the power distribution bus 400.

The depicted components are all controlled by the vehicles energy management system 500, which is connected to vehicle sensors 510.

The vehicle sensors 510 comprise an engine speed sensor, and may also comprise one or more of: an engine torque sensor; a vehicle speed sensor; a battery charge sensor; and a power demand sensor (i.e. a sensor producing data indicative of the user's power demand) When the vehicle is in operation, the energy management system is arranged to determine how to meet the power demand by varying the amount of power to be supplied by the auxiliary power unit 100 and the amount of power to be drawn from or delivered to the battery 200.

In other words, the energy management system 500 controls the hybrid vehicle to determine the flow of power through the power distribution bus 400 between the auxiliary power unit 100, the battery 200 and the load.

Specifically, at times of low load, the energy management system 500 can direct power from the auxiliary power unit to drive the load 300 and charge the battery 200 at the same time.

Similarly, at times of high load, the energy management system 500 can direct power from both the auxiliary power unit 100 and the battery 200 to drive the load 300.

Under braking, the energy management system 500 can direct power generated by the use of the motors 3101, 310r as generators to charge the battery 200. Thus, regenerative braking is possible.

Figure 2 shows a representation of the power circuitry of the hybrid electric vehicle. Figure 3 shows how the power demand of the hybrid electric vehicle can be calculated from the measured voltages and currents on the DC bus.

The following disclosure relates to the determination of optimally fuel efficient power provision by the auxiliary power unit 100 and optimal provision or optimal charging level for the battery 200. Ibis is done by minimising a function that represents the amount of fuel combusted for each unit of energy driving the load 300. The function may include a term that penalises divergence from a preferred level of charge stored in the battery 200 to maintain a desirable state of charge.

In what follows: anu Is the power provided by the auxiliary power unit 100 at time t; Pw)u is the rate of change of P at time t; SoCis the battery State of Charge at time t; the fuel cost In weight of a unit of energy provided by the auxiliary power unit 100; CbWis the fuel cost in weight of a unit of energy provided by the battery 200; and Pba5 the power provided by the battery 200.

For correct operation, the total power demand of the load 300, deri, must be matched by the sum of the power provided by the auxiliary power unit 100 and the power provided by the battery 200, haL' That is delTi = .c1: + ht* A negative hat is possible -this is when the battery 200 draws power from the power distribution bus 500.

Figure 4 shows a graph indicating the efficiency of an internal combustion engine as a function of engine torque and engine speed. Figure 4 shows the optimum effieclency achievable for a given power output as line 700.

The fuel-equivalent cost of power provided by the auxiliary power unit 100 is calculated from efficiency graphs, such as that shown in Figure 4, and the caiorific value of the fuel. Such graphs may be derived by the skilled person using common rrethods.

It is possible to model the cost of providing power with the auxiliary power unit 100 using the following equation: Capu = AtJx6 +kkpu) Qf lapis In this equation: Q is the calorifio value of the fuel in 37kg; is the efficiency of the auxiliary power unit 100 along line 700; k is a transient correction coefficient which determines the influence of the rate of change in the output of auxiliary power unit 100; and is the inverter efficiency. (This may be estimated as 95% unless experimental data is obtained) The second bracketed term in this equation is a transient correction factor intended to model the extra fuel consumption resulting from changing power demands.

Although this is shown as a linear function of the rate of change of P,u, models, which represent a quadratic, or more complicated, relationship are envisaged.

For the second bracketed term a value of k in the range 5% to 10% per 10 kw/s is preferred. More preferably this value is 7% per 10 kw/s.

It is possible to model the cost of charging the battery using the following calculation: I -A bat apu Ci bat chg c',at' So(j = - 1ldiilchg up;, Capuj5 the mean cost of providing the battery 200 with power from the auxiliary power unit 100, measured in grams per second. " may he a fixed value, unchanging for each use of the vehicle. However, the inventors have discovered that tapa can advantageously be measured in a trip-specific manner. In which case, "" can be calculated as the mean cost of providing the battery 200 with power from the auxiliary power unit 100 since the trip commenced.

Like all mean parameters denoted by the bar notation, apu can be calculated by the energy management system 500 using numerical methods for approximate integration such as the trapezium rule. Initiating as zero on each trip allows the energy management system 500 to quickly accumulate an average value that is specific to that trip.

The start and end of a trip way, for example, be indicated by the starting and deactivation ci the vehicle. Alternatively, a reset button (or other input means) may be used to indicate the start of a trip to thereby commence calculation of ThpU It is possible to model the cost of charging the battery using the following equation: SPbc/ ChUt (,ar,SoC) = -cx tlchg 1?cIis apu The average charge cost, , is used to quantify the amount of fuel associated with each unit of energy in the gone into creating each Joule of energy in the battery at the current point in time.

1/djs is the instantaneous discharge efficiency.

In addition to the above-described approaches of using a fixed for every trip, or using a trip-specific cPU it is envisaged that can be calculated as a mean over a predetermined time period ending with the present time.

In other words, the mean value within a trailing window". For example, way be determined as the mean cost of providing power to the battery 200 from the auxiliary power unit 100 over the preceding 5 seconds.

In this way, the system can adapt quickly to driving conditions.

Suitable time periods for such a calculation would be between 30 seconds and 5 minutes.

It is preferred that be calculated using a look-up table or function having engine speed and engine torque as inputs. The table or function can be determined by routine experimentation.

The term is used to describe the portion of total energy in the battery 200 at any time that has come from the auxiliary power unit 100. When the auxiliary power unit 100 charges the battery 200, increases. During a regenerative braking event, decreases, and as a result the fuel-equivalent cost of battery discharge power is reduced. It is determined using: S = 5 I,, dt (ai,api, + a/ree,, For the optimisation of the cost function to also result in usage of power that sustains a target level of charge, it is advantageous to incorporate a term in the cost function that penalises divergence from that target charge.

The battery 200 operates between an upper (SuCH) and lower (SuCL) state of charge recommended value. The energy management system 500 aims to maintain state of charge within those bounds but can venture beyond them if that is the lowest cost course of action.

If the state of charge (SoS) is higher than a desirable level the estimated fuel cost of battery 200 power is reduced. This will cause the cost function to be biased towards using battery 200 power. When state of charge is lower than a desirable level the estimated fuel cost of battery power is increased, thus biasing the cost function towards using auxiliary power unit 100. This is achieved by introducing the scaling term (i-fl) in the following equations: ASoC =21_SoC -JLSOC = SoCH -SoCL KSoCI-J-SoCL) 2 bat = (1 -/3ASOC)C5 (& SoC) In simulation /3=10 was determined to be a reasonable value.

Preferably, the value of SoCL is allowed to vary as a function of vehicle speed. Specifically, a nominal lower allowable charge on the battery 200 is defined, and SoCL is reduced below the nominal lower allowable charge by an amount substantially equal to the kinetic enerqy recoverable under regenerative braking of the vehicle.

The amount of kinetic energy recoverable under regenerative braking is calculated using known methods, taking into consideration the mass of the vehicle, aerodynamic drag and average deceleration.

The amount of kinetic energy recoverable under regenerative braking is substantially propOrtional to the mass of the vehicle. For vehicles that carry a significant load, the mass of the vehicle may vary significantly from the vehicle's unloaded weight. In preferred embodiments, the vehicle's acceleration in response to a known driving force may be measured, and the total mass of the vehicle in combination with its load may be calculated.

The entire fuel-equivalent cost function for charge-sustaining operation can therefore be modelled as: ft (P,, , SoC) = Capy (lDapu, pu) + (1 -/JASOC)Chat (,at, SoC) In other words, the cost function can represent the sum or weighted sum of the cost functions for the auxiliary power unit 100 and the battery 200.

The cost function.J1is the fuel-equivalent cost (the weight in fuel required to provide one unit of energy per second) at time t The energy management system 500 controls the hybrid vehicle to operate in various modes of operation.

In a first mode, at times of high load, both the auxiliary power supply 100 and the battery 200 provide power to the electric motors 3101, 310r via the power distribution bus 400 to meet the power demand of the vehicle.

In this mode, the energy management system 500 must determine how much power is supplied by each of the auxiliary power unit 100 and the battery 200.

This is achieved by first defining a cost function (explained in detail above) that estimates the cost of meeting a total instantaneous power demand of the load 300 when a particular power level is provided by the auxiliary power unit 100. The cost is representative of the weight of fuel combusted for each unit of energy per second delivered to the load.

As explained above, the cost function does not only consider the efficiency in the generation of electrical power by the combustion of fuel in the internal combustion engine 150, but also the cost of supplementing that power with power from the battery 200 in order to meet the total power demand of the load 300.

It is not always optimal to simply provide the instantaneous power demand of the load 300 with the auxiliary power unit 100 because transient power demands introduce further inefficiencies into the supply of power by an engine 150. That is, the cost of varying the power supplied by the auxiliary power unit 100 to closely follow a fast changing demand over a particular time period is greater than the cost of consistently providing the average power demand over that period.

The energy management system removes the dependency of the auxiliary power unit 100 on the instantaneous power demand and makes up the difference by providing power from the battery 200 (or by charging the battery 200) In order to remove the dependence of the auxiliary power unit 100 upon the instantaneous power demand of the load 300, the energy management system evaluates the cost function on a plurality of trial values of power supplied by the auxiliary power unit 100.

The cost function therefore provides an estimate of the cost of meeting the instantaneous demand using both the battery 200 and the auxiliary power unit 100 for each of the trial values of power supplied by the auxiliary power unit 100.

Each of the plurality of trial values of power supplied by the auxiliary power unit 100 are determined as the average power demand over a preceding time period (where the time periods differ in duration) -14-For example, a first trial value might be the average power demand of the load 300 over the previous 10 seconds, while a second trial value might be the average power demand of the load 300 over the previous 5 seconds, and a third trial value might be the average power demand of the load 300 over the previous 1 second.

The cost function is evaluated for each of the trEal values. The energy management system 500 selects the trial value that results in the lowest cost in terms of weight of fuel combusted to deliver each unit of energy.

This trial value represents the power delivered by the auxiliary power unit 100 with the difference between that power and the instantaneous power demand being met by the battery 200.

The number of trial values is limited only by the rate of calculation of the energy management system 500.

Preferably, the energy management system will calculate the cost of a plurality of trial values periodically.

Preferably, the calculation will occur every 0.01 seconds to every 1 second. Most preferably, the calculatEon will occur every 0.1 seconds.

Preferably, there will be 200 trial values calculated as the average power demand over preceding time period.

Each trial value has a different associated time period.

Preferably, the time periods have lengths up to 20 seconds.

Since the time periods all end at the present time, a trial value relating to 0 seconds correspond to the instantaneous demand. Preferably, this instantaneous demand will be included as one of the trial values.

For example, the trial values may be calculated for twenty-one time periods having durations of 0 seconds, 1 second, 2 seconds,... 20 seconds.

In this way, the energy management system 500 can determine a desired power to be delivered by the auxiliary power unit 100 as the average power demand that results in the lowest value of the power cost function.

In a second mode of operation of the hybrid vehice, when the load is not high, only the auxiliary power supply 100 provides power to the electric motors 3101, 310r via the power distribution bus 400 to meet the power demand of the vehicle. In this mode, the auxiliary power unit 100 can also provide power to the power distribution bus 500 to charge the battery 200.

In the second mode, the auxiliary power unit 100 provides power to both the load 300 to drive the vehicle and to the battery 200 to increase its level of stored charge.

The energy management system 500 must determine the most efficient amount of power to be provided by the auxiliary power unit 100 and the amount of power supplied to charge the battery 200.

In the second mode of operation, the cost function is minimised in the same way as in the first mode. The difference between the first and second modes of operations is that in the first mode aL is positEve, and in the second mode ha-is negative. That is, in the second mode the amount of power delivered by the auxiliary power unit 100 is equal to the sum of the instantaneous power demand and the power oharging the battery 200.

A third mode of operation is used when the vehiole is braking. In this mode, the energy management system 500 direots electrical power generated by the electric motors 3101, 31Cr (driven by the rotation of the wheels 3401, 34Cr under braking) via the power distribution bus 400 to recharge the battery 200.

In addition to the above described modes of operation followed by the energy management system 500, a number of control rules are preferably followed in order to ensure that the battery state of charge can be maintained within desirable limits.

When the power demand of the load 300 exceeds a predefined maximum level of power output for the auxiliary power unit 100, the auxiliary power unit is set to operate at that predefined maximum level.

When the charge on the battery 200 drops below a target level, trial values of P-, are excluded from evaluation.

In other words the power cost function is only evaluated for each of the average power demands calculated for each time period if those average power demands are greater than or equal to the instantaneous power demand PJ,* The target level for the charge on the battery may be in the range of 5% to 15%, and preferably approximately 10%, above the predetermined level below which the physical properties of the battery will be degraded.

When the charge cn the battery 200 drops below a target level, the trial values of used as an input to the cost function are artificially increased. That is, the cost function estimates the cost of providing a power level equal to the product of the trial values calculated as moving averages and a load scaling factor.

When the charge on the battery 200 drops below a target level, the load scaling factor is increased in dependence upon the difference between the charge on the battery 200 and the target level.

Preferably, when the charge on the battery 200 drops below a target level, the load scaling factor is increased by Z% for every 1% difference between the target level and the battery 200 charge, e.g. if the battery state of charge is 43% and the target state of charge is 50%, then the load scaling factor is 1 -0.01(7 x Z) . Experimental results show a value of Z of 1.1 is a suitable value. Thus in the example given above, the load scaling factor would be 1.07, i.e. the trial value of power demand would be 107% of the original value.

Whereas the cost function set out above calculates the cost of providing power in terms of amount of fuel per unit of energy delivered, it is envisaged that other costs may be calculated, such as the cost in terms of the amount of undesirable exhaust emissions per unit of energy delivered (e.g. weight in NO output per kW energy delivered) Figure 5 shows a flow chart depicting a preferred embodiment of a method of operating a series hybrid vehicle.

In step 800, the energy management system compares the current state of charge of the battery 200 with a threshold charge (in this case 40%) If the state of charge has not dropped below the threshold charge since the battery was last charged by an external power supply then in step 805, the battery 200 is used as the sole source of power.

If the state of charge has dropped below the threshold charge since the battery was last charged by an external power supply then in step 810, the engine switches to a charge-sustaining mode of operation.

In step 815, the energy management system measures the instantaneous power demand from the DC bus.

In step 820, the energy management system calculates the average power demand over each of a plurality of time periods of differing lengths and ending with the present time. These are used as trial values. Preferably the instantaneous power demand is also used as a trial value.

In step 825, the energy management system compares the current state of charge of the battery 200 with a preferred charge.

If the state of charge is below the preferred charge then in step 830, the total power demand of the engine is determined as the product of a scaling factor with the instantaneous power demand. Preferably, the total power demand is increased above the measured instantaneous power demand by 1.1% for each 1% that the state of charge is below the preferred oharge.

In step 835, the energy management system checks whether the total power demand is greater than that achievable by the auxiliary power unit 100. If so, then in step 845, the auxiliary power unit 100 is instructed to provide its maximum power output.

If, in step 825 the state of charge is not below the preferred obarge then in step 840, or if the total power demand is below the maximum power achievable by the auxiliary power unit 100, the cost function is evaluated for each of the trial P.1 values and the auxiliary power unit 100 is instructed to provide the power output corresponding to the UU value with the lowest associated cost.

In preferred embodiments the battery cost function may include a term representing the cost of charging the battery 200 using an external power supply. Such embodiments can provide optimisation of the cost function with respect to both the power provided both by the auxiliary power unit 100 and also the external power supply.

In preferred embodiments the battery cost function may include a term representing the temperature of the battery 200, so that the battery cost function increases the cost of providing power with the battery 200 when the temperature is above a threshold value. This can allow -20 -the cost function to be used to prevent damage to the battery.

In preferred embodiments the battery cost function may include a term that increases the battery cost function when the state of charge of the battery 200 exceeds an upper limit or drops below a lower limit. This can allow the cost function to be used to prevent damage to the battery.

Whereas in the description given above, the APU cost function and the battery cost function represent the amount of fuel per unit of energy delivered, this is merely a preferred cost metric on which to base the distribution of power.

It is envisaged that for different applications, it may be preferred to base the cost metric upon: the cost in monetary terms per unit of energy delivered; the oost in terms of amount of undesirable emissions (e.g. CO2.

NOR, etc.) produced per unit of energy delivered; the cost In terms of fatigue of one or more components of the auxiliary power unit; or the cost In terms of reduction in the life of the battery.

Indeed, it is envisaged that a weighted combination of the above-listed costs may provide a suitable cost metric. -21 -

Claims

CLAIMS: 1. A series hybrid vehicle comprising: an energy management system; a battery, an auxiliary power unit; a ioad comprising an electric motor; and an electrical power bus via which electrical power can be delivered between the auxiliary power unit, the battery and the load, wherein the series hybrid electric vehicle has a first operating mode in which: the energy management system determines a target power demand; the energy management system calculates an APU cost function representing the cost of supplying power from the auxiliary power unit; the energy management system calculates a battery cost function representing the cost of supplying power from the battery; the energy management system calculates a desired APU power to be supplied by the auxiliary power unit and a desired battery power to be supplied by the battery so as to minimise the weighted sum of the APtJ cost function and the battery cost function; and the load is supplied with the desired APU power from the auxiliary power unit and the desired battery power from the battery power so that the total power delivered to the load is equal to the determined target power demand; and wherein the weighted sum of the APU cost function and the battery cost function represents one of, or a weighted combination of: the amount of fuel per unit of energy delivered; -22 -the cost in monetary terms per unit of energy delivered; the cost in terms of amount of undesirable emissions produced per unit of energy delivered; the cost in terms of fatigue of one or more components of the auxiliary power unit; or the cost in terms of reduction in the life of the battery.
2. The series hybrid vehicle of claim 1, wherein the battery cost function includes a term representing the cost of charging the battery using an external power supply thereby providing optimisation with respect to both the power provided by the auxiliary power supply and also by the external power supply.
3. A series hybrid vehicle comprising: an energy management system; a battery, an auxiliary power unit; a load comprising an electric motor; and an electrical power bus via which electrical power can be delivered between the auxiliary power unit, the battery and the load, wherein the series hybrid electric vehicle has a first operating mode in which: the energy management system determines a target power demand; the energy management system calculates an APU cost function representing the cost of supplying power from the auxiliary power unit; the energy management system calculates a battery cost function representing the cost of supplying power from the battery; -23 -the energy management system calculates a desired APU power to be supplied by the auxiliary power unit and a desired battery power to be supplied by the battery so as to minimise the weighted sum of the APtJ cost function and the battery cost function; and the load is supplied with the desired APU power from the auxiliary power unit and the desired battery power from the battery power so that the total power delivered to the load is equal to the determined target power demand, and wherein the target power demand is determined by measuring the electrical demand on the electrical power bus.
4. The series hybrid vehicle of claim 3, wherein the power demand is determined solely in dependence upon the electrical demand on the electrical power bus.
5. A series hybrid vehicle comprising: an energy management system; a battery, an auxiliary power unit; a load comprising an electric motor; and an electrical power bus via which electrical power can be delivered between the auxiliary power unit, the battery and the load, wherein the series hybrid electric vehicle has a first operating mode in which: the energy management system determines a target power demand; the energy management system calculates an APU cost function representing the cost of supplying power from the auxiliary power unit; -24 -the energy management system calculates a battery cost function representing the cost of supplying power from the battery; the energy management system calculates a desired APU power to be supplied by the auxiliary power unit and a desired battery power to be supplied by the battery so as to minimise the weighted sum of the APU cost function and the battery cost function; and the load is supplied with the desired APU power from the auxiliary power unit and the desired battery power from the battery power so that the total power delivered to the load is equal to the determined target power demand, and wherein: the battery cost function includes a battery charging term that varies in proportion to the cost of charging the battery with the auxiliary power unit; and the battery charging term is calculated by fInding the mean cost of supplying power from the auxiliary power unit to the battery over a time period ending with the present time.
6. The series hybrid of claim 5, wherein the battery charging term is calculated by finding the mean cost of supplying power from the auxiliary power unit to the battery over a time period starting with the beginning of a trip and ending with the present time.
7. The series hybrid of claim 5, wherein the battery charging term is calculated by finding the mean cost of supplying power from the auxiliary power unit to the battery over a time period of predetermined duratIon ending with the present time.

-25 -
8. The series hybrid of any one of claims 5 to 7, wherein cost of supplying power from the auxiliary power unit to the battery is calculated from a look-up table having engine speed and engine load as inputs.
9. A series hybrid vehicle comprising: an energy management system; a battery, an auxiliary power unit; a load comprising an electric motor; and an electrical power bus via which electrical power can be delivered between the auxiliary power unit, the battery and the load, wherein the series hybrid electric vehicle has a first operating mode in which: the energy management system determines a target power demand; the energy management system calculates an APU cost function representing the cost of supplying power from the auxiliary power unit; the energy management system calculates a battery cost function representing the cost of supplying power from the battery; the energy management system calculates a desired APU power to be supplied by the auxiliary power unit and a desired battery power to be supplied by the battery so as to minimise the weighted sum of the APtJ cost function and the battery cost function; and the load is supplied with the desired APU power from the auxiliary power unit and the desired battery power from the battery power so that the total power delivered to the load is egual to the determined target power demand, and -26 -wherein the energy management system calculates a desired APU power by: calculating an average power demand for each of a plurality of time periods of different lengths; evaluating the power cost function for each of the average power demands calculated for each time period; and setting a desired APU power as the average power demand that results in the lowest value of the power cost function.
10. The series hybrid vehicle of claim 9, wherein the/each time period ends with and includes the present time.
11. The series hybrid vehicle of claim 9 or 10, wherein the energy management system evaluates the power cost function for each of the average power demands calculated for each time period only if the average power is greater than or equal to the target power demand.
12. A series hybrid vehicle comprising: an energy management system; a battery, an auxiliary power unit; a load comprising an electric motor; and an electrical power bus via which electrical power can be delivered between the auxiliary power unit, the battery and the load, wherein the series hybrid electric vehicle has a first operating mode in which: the energy management system determines a target power demand; -27 -the energy management system calculates an APU cost function representing the oost of supplying power from the auxiliary power unit; the energy management system oalculates a battery oost function representing the cost of supplying power from the battery; the energy management system oalculates a desired APU power to be supplied by the auxiliary power unit and a desired battery power to be supplied by the battery so as to minimise the weighted sum of the APtJ cost function and the battery cost function; and the load is supplied with the desired APU power from the auxiliary power unit and the desired battery power from the battery power so that the total power delivered to the load is egual to the determined target power demand; and wherein: the energy management system prevents the charge on the battery from dropping below a minimum allowable battery charge; the minimum allowable battery charge is reduced from a fixed minimum battery charge in dependence upon vehicle speed by an amount egual to the amount of charge recoverable under braking of the vehicle.
13. The series hybrid vehicle of claim 12, whereTh the amount of charge recoverable under braking the vehicle is determined based on the vehicle's mass.
14. A series hybrid vehicle comprising: an energy management system; a battery, an auxiliary power unit; a load comprising an electric motor; and -28 -an electrical power bus via which electrical power can be delivered between the auxiliary power unit, the battery and the load, wherein the series hybrid electric vehicle has a first operating mode in which: the energy management system determines a target power demand; the energy management system calculates an APU cost function representing the cost of supplying power from the auxiliary power unit; the energy management system calculates a battery cost function representing the cost of supplying power from the battery; the energy management system calculates a desired APU power to be supplied by the auxiliary power unit and a desired battery power to be supplied by the battery so as to minimise the weighted sum of the APU cost function and the battery cost function; and the load is supplied with the desired APU power from the auxiliary power unit and the desired battery power from the battery power so that the total power delivered to the load is egual to the determined target power demand; and wherein: the energy management system monitors the actual power demand of the load and the present charge on the battery; the energy management system compares the present charge on the battery with a target battery charge; if the target battery charge is not lower than the present charge on the battery, the energy management system sets the target power demand as product of the actual power demand and a scaling factor that increases -29 -in dependence upon the difference between the target battery charge and the present charge cn the battery.
15. The series hybrid vehicle of claim 14, wherein if the target battery charge is lower than the present charge on the battery, the energy management system sets the target power demand as the actual power demand of the load -
16. A series hybrid vehicle comprising: an energy management system; a battery, an auxiliary power unit; a load comprising an electric motor; and an electrical power bus via which electrical power can be delivered between the auxiliary power unit, the battery and the load, wherein the series hybrid electric vehicle has a first operating mode in which: the energy management system determines a target power demand; the energy management system calculates an APU cost function representing the cost of supplying power from the auxiliary power unit; the energy management system calculates a battery cost function representing the cost of supplying power from the battery; the energy management system calculates a desired APU power to be supplied by the auxiliary power unit and a desired battery power to be supplied by the battery so as to minimise the weighted sum of the APU cost function and the battery cost function; and the load is supplied with the desired APU power from the auxiliary power unit and the desired battery power -30 -from the battery power so that the total power delivered to the load is equal to the determined target power demand; and wherein: the battery cost function includes a charge source term representing the proportion of the present charge on the battery that has been provided by the auxiliary power unit.
17. The series hybrid vehicle of ciaim 14, wherein the charge source term represents the proportion of the present charge on the battery that has been provided by the auxiliary power unit over a single trip.
18. The series hybrid vehicle of claim 14, wherein the charge source term representing the proportion of the present charge on the battery that has been provided by the auxiliary power unit over the lifetime of the battery.
19. The series hybrid vehicle of any preceding claim, wherein if the desired APU power is greater than a maximum APU power, then setting the desired APU power to the maximum APU power.
20. The series hybrid vehicle of any preceding claim, wherein the sum of the APU cost function and the battery cost function is weighted at least in part in dependence upon the difference between the present charge on the battery and the desired battery charge.
21. The series hybrid vehicle of any preceding claim, wherein the APU cost function includes a term that varies -31 -in dependence upon the associated rate of change of power provided by the auxiliary power unit.