Predicting energy consumption for an electric vehicle using fluctuations in past energy consumption

Abstract

Procedures for controlling an electric vehicle, which include the following: Changing the operation of an electric vehicle in response to a predicted energy consumption value; characterized in that the method further comprises the following: Adjusting the predicted energy consumption in response to a standard deviation of fluctuations in past energy consumption values ​​(R); and Reducing a predicted remaining range (DTE) in a non-linear manner using the adjusted predicted energy consumption as a battery state of charge (BSOC, CSOC) decreases, where the predicted energy consumption value X is calculated using the following equation (1): X = μ+ z'* σ , where µ is the mean energy consumption value during a driving cycle, σ is the standard deviation of the energy consumption value during the driving cycle, and z' is a weighting factor of σ.

Description

[0001] This disclosure relates to the prediction of the energy consumption of electric vehicles and, in particular, to the use of fluctuations in past energy consumption to improve the prediction. The prediction of the energy consumption of an electric vehicle is disclosed, for example, in DE 10 2012 216 115A1, DE 11 2006 000 895 T5, DE 60 2004 010 203 T2, DE 10 2012 206 410 A1, DE 10 2012 213 320 A1, and US 2013 / 0 166 123 A1.

[0002] In general, electric vehicles differ from conventional motor vehicles in that electric vehicles are powered by one or more battery-powered electric motors. Conventional motor vehicles, on the other hand, rely exclusively on an internal combustion engine to propel the vehicle. Electric vehicles can use electric motors instead of, or in addition to, the internal combustion engine.

[0003] Examples of electric vehicles include hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles (PHEVs), fuel cell vehicles, and battery electric vehicles (BEVs). An electric vehicle powertrain typically includes a battery pack containing battery cells that store electrical energy to power the electric motor. The battery cells can be charged before use. They can also be recharged while driving through regenerative braking or an internal combustion engine.

[0004] Predicting an electric vehicle's energy consumption is useful, for example, to estimate the distance the battery will travel. Some vehicles predict energy consumption solely based on an average consumption over specific time periods, such as the last five, fifteen, or thirty miles driven, or the last five, fifteen, or hour of driving time. The customer can choose between mileage and driving time.

[0005] According to the invention, a method according to claim 1, a method according to claim 9, and an electric vehicle according to claim 10 are provided.

[0006] A method for controlling an electric vehicle according to an exemplary aspect of the present disclosure includes, among other things, changing the operation of an electric vehicle depending on a predicted energy consumption value. The method also includes adjusting the predicted energy consumption in response to fluctuations in past energy consumption values.

[0007] In another example of the preceding procedure, the process involves adjusting the predicted energy consumption value in response to average values ​​of past energy consumption values.

[0008] In another example of one of the preceding procedures, the process involves changing the weighting of fluctuations in past energy consumption values ​​during the adjustment.

[0009] In another example of one of the preceding methods, the procedure involves increasing the weighting of fluctuations in past energy consumption values ​​based on a reduced state of charge.

[0010] In another example of one of the preceding methods, the procedure involves reducing the weighting of the average values ​​of past energy consumption based on a reduced state of charge.

[0011] In another example of one of the preceding methods, the past energy consumption values ​​are energy consumption values ​​for the vehicle for a range of distances driven, times, or both.

[0012] In another example of one of the preceding methods, the area can be adjusted in response to a command from an operator of the electric vehicle.

[0013] In another example of one of the preceding procedures, the fluctuations in past energy consumption values ​​comprise a standard deviation of past energy consumption values.

[0014] An electric vehicle according to another exemplary aspect of the present disclosure includes, among other things, a data storage module for holding a range of past energy consumption values ​​for an electric vehicle and a controller for calculating a predicted energy consumption for the electric vehicle based on fluctuations in that range.

[0015] In another example of the previous electric vehicle, the fluctuations in the range comprise one standard deviation of the range.

[0016] In another example, using one of the previous electric vehicles, the controller calculates the predicted energy consumption for the electric vehicle based on average values ​​of past energy consumption in the area.

[0017] According to yet another exemplary aspect of the present disclosure, an electric vehicle includes, among other things, a data storage module for holding a range of past energy consumption values ​​for the electric vehicle. A controller is provided to calculate a predicted energy consumption for the electric vehicle based on fluctuations in this range.

[0018] The various features and advantages of the revealed examples become clear to experts in this field through the detailed description. The figures accompanying the detailed description can be briefly described as follows: Fig. Figure 1 schematically shows an example electric vehicle powertrain. Fig. Figure 2 graphically illustrates the energy consumption during an exemplary driving cycle of a vehicle with the powertrain of Fig. 1. Fig. Figure 3 graphically illustrates the energy consumption for another exemplary driving cycle of the vehicle with the powertrain of Fig. 1. Fig. Figure 4 shows a highly schematic view of an exemplary electric vehicle. Fig. Figure 5 schematically shows the process of an exemplary procedure for predicting an energy consumption value for the electric vehicle of Fig. 4. Fig. Figure 6 shows graphical examples of weightings for fluctuations in energy consumption values ​​at different battery charge levels when energy consumption is measured using the method of Fig. 5 is predicted. Fig. Figure 7 shows exemplary coverage of the predicted energy consumption values ​​for different weightings of fluctuations. Fig. Figure 8 graphically illustrates how a remaining range prediction that uses fluctuations in past energy consumption can differ from a remaining range prediction that does not use fluctuations in past energy consumption.

[0019] Fig. Figure 1 schematically shows a powertrain 10 for an electric vehicle. Although it is depicted as a hybrid electric vehicle (HEV), it is understood that the concepts described here are not limited to HEVs and could extend to other electrified vehicles, including but not limited to plug-in hybrid electric vehicles (PHEVs), fuel cell vehicles, and battery electric vehicles (BEVs).

[0020] In one embodiment, the powertrain 10 is a power-split powertrain system that uses a first drive system and a second drive system. The first drive system comprises a combination of a power machine 14 and a generator 18 (i.e., a first electric machine). The second drive system comprises at least one motor 22 (i.e., a second electric machine), the generator 18, and a battery pack 24. In this example, the second drive system is considered to be an electric drive system of the powertrain 10. The first and second drive systems generate torque to drive one or more sets of vehicle drive wheels 28 of the electric vehicle.

[0021] The power engine 14, which in this example is an internal combustion engine, and the generator 18 can be connected via a power transmission unit 30, such as a planetary gear set. Of course, other types of power transmission units, including other gear sets and transmissions, can also be used to connect the power engine 14 to the generator 18. In a non-limiting embodiment, the power transmission unit 30 is a planetary gear set comprising a ring gear 32, a sun gear 34, and a carrier assembly 36.

[0022] The generator 18 can be driven by the power machine 14 via the power transmission unit 30 to convert kinetic energy into electrical energy. Alternatively, the generator 18 can function as a motor to convert electrical energy into kinetic energy and thus deliver torque to a shaft 38 connected to the power transmission unit 30. Because the generator 18 is operationally linked to the power machine 14, the rotational speed of the power machine 14 can be controlled by the generator 18.

[0023] The ring gear 32 of the power transmission unit 30 can be connected to a shaft 40, which is connected to the vehicle drive wheels 28 by a second power transmission unit 44. The second power transmission unit 44 can comprise a gear set with multiple gears 46. Other power transmission units may also be suitable. The gears 46 transmit torque from the power unit 14 to a differential 48 to ultimately provide tractive force for the vehicle drive wheels 28. The differential 48 can comprise multiple gears that enable the transmission of torque to the vehicle drive wheels 28. In this example, the second power transmission unit 44 is mechanically coupled to an axle 50 through the differential 48 to distribute the torque to the vehicle drive wheels 28.

[0024] The motor 22 (i.e., the second electric machine) can also be used to drive the vehicle's drive wheels 28 by delivering torque to a shaft 52, which is also connected to the second power transmission unit 44. In one embodiment, the motor 22 and the generator 18 operate together as part of a regenerative braking system, in which both the motor 22 and the generator 18 can be used as motors to deliver torque. For example, the motor 22 and the generator 18 can each deliver electrical power to the battery pack 24.

[0025] The battery pack 24 is an example type of electric vehicle battery assembly. The battery pack 24 can take the form of a high-voltage battery that can supply electrical power to operate the motor 22 and the generator 18. Other types of energy storage devices and / or output devices can also be used with the electric vehicle, which has the powertrain 10.

[0026] The energy consumption value for a vehicle with powertrain 10 changes when the vehicle is being propelled. Energy consumption values ​​vary depending on variables that affect energy consumption. Examples of factors influencing energy consumption include, but are not limited to, the way the vehicle accelerates, the way the vehicle is stopped, road gradients, road conditions, driving environments, and vehicle accessories that consume energy.

[0027] Past energy consumption data is frequently used to provide an expected or predicted energy consumption value for the vehicle. The predicted energy consumption value can be used in many ways. For example, it can be used to estimate a vehicle's remaining range (DTE). Generally, the DTE estimate is equal to the available energy divided by the predicted energy consumption value. Vehicle operators rely on the DTE estimate, among other things, to estimate whether they can drive the vehicle to a desired location without recharging or accessing the power of an internal combustion engine.

[0028] With reference to the Fig. 2 and Fig. Figure 3 shows a graph 100 representing an energy consumption value R for a first driving cycle and a graph 200 representing an energy consumption value R for a second driving cycle.

[0029] An average (or mean) energy consumption value A1 for the first driving cycle for a distance X is equal to or approximately equal to an average energy consumption value A2 for the second driving cycle for the distance X.

[0030] Remarkably, the energy consumption value for the first driving cycle varies or fluctuates more than the energy consumption value for the second driving cycle. The average values ​​A1 and A2 do not take these fluctuations between the driving cycles into account. If average energy consumption values ​​were used to calculate a predicted energy consumption value without considering these fluctuations, the predicted energy consumption value for the first driving cycle would be the same as the predicted energy consumption value for the second driving cycle.

[0031] In this example, the predicted energy consumption value for the vehicle is adjusted in response to fluctuations in the driving cycles. The predicted energy consumption value based on the first driving cycle in graph 100 differs from a predicted energy consumption value based on the second driving cycle in graph 100, even though the mean values ​​A1 and A2 are the same.

[0032] Fluctuations in energy consumption values ​​for the first and second driving cycles can be quantified using the standard deviation. These fluctuations indicate the extent of the deviation from the respective mean values ​​A1 and A2 in the first and second driving cycles. In this example, graph 100 would show a higher standard deviation than graph 200.

[0033] With reference to Fig. 4 and Fig. Figure 5 comprises an exemplary electric vehicle 300, a data storage module 304, a control unit 308, and a battery 316, which are operationally interconnected. The data storage module 304 contains the past energy consumption values ​​for the driving cycles of the electric vehicle 300, such as those graphically represented in the Fig. 2 and Fig. 3 data shown. The data storage module 304 and the control unit 308 can be arranged internally in the vehicle 300, externally from the vehicle 300, or both ways.

[0034] The control unit 308 includes a processor 316 to execute a program that predicts the energy consumption for the vehicle 300. This prediction can be used, for example, to provide a DTE estimate for the vehicle 300. The predicted energy consumption is based, at least in part, on fluctuations within a range of past energy consumption values ​​for the vehicle 300.

[0035] The processor 316 is programmed to execute some or all of the steps in a procedure 320 for controlling the electric vehicle 300. The exemplary procedure 320 includes a step 326 for calculating the fluctuations in the past energy consumption values ​​for the vehicle 300. In a step 330, the procedure 320 adjusts a predicted energy consumption for the vehicle 300 in response to fluctuations in past energy consumption rates. In a step 334, the procedure 320 modifies the operation of the vehicle 300 in response to the adjustment.

[0036] In step 330, the predicted energy consumption value can further be adjusted in response to mean values ​​of past energy consumption values ​​or other information and variables. In some examples, procedure 320 can change the weighting of the fluctuations of past energy consumption values ​​relative to the weighting of the mean values ​​of past energy consumption values, depending on, for example, the charge level in a battery 312 of the vehicle 300. Increasing the weighting of the fluctuations relative to the weighting of the mean values ​​increases the influence of the fluctuations on the predicted energy consumption value.

[0037] In step 334, the modification can include the vehicle taking a different route, driving less aggressively, extending a planned route, switching off certain systems to save energy, etc. For example, if the predicted energy consumption from step 330 indicates that a desired destination cannot be reached without recharging battery 312, step 334 can suggest taking a different route to reach a charging station before the desired destination.

[0038] In one example, a predicted energy consumption value X of the first driving cycle is calculated using the following equation (1): X=μ+z'*σ.

[0039] In equation (1), µ is the mean energy consumption value during the driving cycle, σ is the standard deviation of the energy consumption value during the driving cycle, and z' is a weighting factor of σ. The predicted energy consumption value for vehicle 300 is then used to estimate the DTE.

[0040] In some examples, a vehicle operator can adjust the rate at which changes in µ and σ are detected and how they are used to calculate the predicted energy consumption value and estimate the DTE. For example, the operator can adjust the detection rate to update the data every five, fifteen, or fifty miles driven by the vehicle.

[0041] When significant fluctuations in past driving cycle energy consumption are included, the energy consumption forecast is often more conservative. For example, a DTE calculated using energy consumption forecasted with fluctuations might be 100 miles for a nominal charge of battery pack 24. In contrast, a DTE calculated using energy consumption forecasted without fluctuations might be 95 miles for the nominal charge percentage of battery pack 24.

[0042] The value of the weighting factor z' is calibratable and can be set based on whether a more or less conservative forecast of energy consumption is desired. As in Fig. As shown in Figure 6, the value of z' can vary based on the state of charge of battery pack 24. The value of z' can be a function of the state of charge of battery pack 24. The value of z' can adjust automatically, in response to operator input, or both.

[0043] In this example, the DTE is only conservative in the lower range of the charge level. In the lower charge ranges, there is often an increased potential for range anxiety. The more conservative DTE helps to alleviate this range anxiety.

[0044] Depending on the state of charge, the DTE calculation may need to consider different coverage levels. Coverage represents the expected probability of trips where the number of miles a customer drives before a full charge is greater than or equal to the displayed DTE figure. The value of z' can be adjusted to ensure these coverage levels are met.

[0045] With regard to the distribution of Fig. 7. For example, when the customer-usable state of charge (CSOC) is between seventy and one hundred percent, the DTE estimate may require fifty percent coverage, as shown in 400. When the CSOC is reduced to approximately twenty percent, the DTE estimate may require eighty-four percent coverage, as shown in 404. When the CSOC is reduced to approximately zero percent, the DTE estimate may require approximately ninety-eight percent coverage, as shown in 408.

[0046] In some examples, the vehicle operator can adjust the value of z' as desired. For instance, the operator can choose between a more conservative DTE or a less conservative DTE. The more conservative DTE would be calculated using a predicted energy consumption with a higher weighting value z' than the less conservative DTE.

[0047] The exemplary approach to calculating the DTE with fluctuations in past energy consumption can help to extend the usable charging range of battery pack 24 (i.e., the CSOC range). Referring to Fig.Section 8 provides an example for the CSOC range of P1 to P3, representing a percentage of the total battery state of charge (BSOC). This is the CSOC used to calculate the DTE with an adjustment to account for fluctuations in past energy consumption. Alternatively, a CSOC range of P1 to P2 is used to calculate the DTE without any adjustment to account for fluctuations in past energy consumption.

[0048] DTEwith_var represents the DTE across the CSOC range when the DTE is adjusted in response to fluctuations in past energy consumption values. DTEno_var represents the DTE across the CSOC range when the DTE is calculated without adjustment in response to fluctuations in past energy consumption values.

[0049] The state-of-charge (SOC) range from P2 to P3 percent can be used in the restricted operating strategy (LOS). This portion of the state-of-charge value is not included when calculating a DTEno_var. Instead, the DTE remains zero even when the total battery state-of-charge capacity is still at P2 percent, to ensure that some range remains when DTEno_var is zero.

[0050] The DTEwith_var example is more conservative (and reliable). Therefore, more coverage is available with a lower BSOC. The LOS portion of the CSOC range can thus be included when calculating DTEwith_var. By including fluctuations when calculating past energy consumption, the BSOC in this example can be reduced to less than P2 percent. Similarly, the new DTE calculation can result in a higher DTE when the battery pack is fully charged, as the lower limit of the CSOC is reduced to less than P2 percent of the BSOC.

[0051] In some examples, the CSOC ranges from P1 percent to P3 percent of the total BSOC. The predicted energy consumption value X increases as the available energy in battery pack 24 decreases. The example DTE begins to decrease non-linearly (at 30 percent BSOC in this example) as the predicted energy consumption value X increases and begins to be based on the standard deviation of the energy consumption value during the driving cycle. The rate at which the DTE changes decreases as the CSOC approaches zero percent. At zero percent CSOC, the DTE is zero miles of range.

[0052] Features of at least some of the disclosed examples include a more conservative DTE, which can reduce an operator's range anxiety. The initial DTE is also increased in some examples.

Claims

[1] Method for controlling an electric vehicle comprising: Changing the operation of an electric vehicle in response to a predicted energy consumption value; characterized by , that the procedure further includes the following: Adjusting the predicted energy consumption in response to a standard deviation of fluctuations in past energy consumption values ​​(R); and Reducing a predicted remaining range (DTE) in a non-linear manner using the adjusted predicted energy consumption as a battery state of charge (BSOC, CSOC) decreases, where the predicted energy consumption value X is calculated using the following equation (1): X=μ+z'*σ, where µ is the mean energy consumption value during a driving cycle, σ is the standard deviation of the energy consumption value during the driving cycle, and z' is a weighting factor of σ. [2] The method of claim 1, further comprising adjusting the predicted energy consumption in response to mean values ​​(A1, A2) of past energy consumption values ​​(R). [3] The method of claim 2, further comprising changing a weighting of fluctuations in past energy consumption values ​​(R) during the adjustment. [4] Method according to claim 3, further comprising increasing the weighting of fluctuations in past energy consumption values ​​(R) based on a reduced state of charge (BSOC, CSOC). [5] The method of claim 3, further comprising reducing the weighting of the mean values ​​of the past energy consumption values ​​(R) based on a reduced state of charge (BSOC, CSOC). [6] Method according to claim 2, wherein the past energy consumption values ​​(R) are energy consumption values ​​for the vehicle for a range of distances driven. [7] Method according to claim 6, wherein the area can be adjusted in response to a command from an operator of the electric vehicle. [8] The method of claim 1, further comprising extending a range of a customer-usable state of charge (CSOC) in response to the adjustment. [9] Method (320) for controlling an electric vehicle, comprising the following: Changing (334) the operation of an electric vehicle in response to a predicted energy consumption value; characterized by , that the procedure further includes the following: Adjusting (330) the predicted energy consumption in response to a standard deviation of fluctuations in past energy consumption values; where the change involves selecting a route to drive based on the predicted energy consumption value, where the predicted energy consumption value X is calculated using the following equation (1): X=μ+z'*σ, where µ is the mean energy consumption value during a driving cycle, σ is the standard deviation of the energy consumption value during the driving cycle, and z' is a weighting factor of σ. [10] Electric vehicle (300) comprising the following: a data storage module (304) that holds a range of past energy consumption values ​​(R) for the electric vehicle (300); characterized by , that the electric vehicle (300) further includes the following: a controller (308) to calculate a predicted energy consumption for the electric vehicle (300) based on a standard deviation of fluctuations in the range, wherein the controller (308) non-linearly reduces a predicted remaining range (DTE) using the adapted predicted energy consumption as a battery state of charge (BSOC, CSOC) decreases, and The controller (308) calculates the predicted energy consumption value X using the following equation (1): X=μ+z'*σ, where µ is the mean energy consumption value during a driving cycle, σ is the standard deviation of the energy consumption value during the driving cycle, and z' is a weighting factor of σ. [11] Electric vehicle (300) according to claim 10, wherein the controller further calculates the predicted energy consumption for the electric vehicle (300) based on mean values ​​(A1, A2) of the past energy consumption values ​​(R) in the range. [12] Electric vehicle (300) according to claim 11, wherein the controller weights the fluctuations and the mean values ​​(A1, A2) differently during the calculation depending on a battery charge state (BSOC, CSOC) of the electric vehicle (300). [13] Electric vehicle (300) according to claim 11, further comprising a weighting selector module to receive an input that causes the controller to weight the fluctuations and the mean values ​​(A1, A2) differently during the calculation. [14] Electric vehicle (300) according to claim 11, further comprising a range selector to receive an input that adjusts the size of the range.

Images