Method and system for internal combustion engine and powertrain control

Abstract

Procedure comprising: adjusting data points of a vehicle powertrain calibration table using data collected on-board a vehicle (100), and using data downloaded from an off-board network (13), wherein the downloaded data is collected on-board one or more other vehicles (100) communicating with the network (13), determining whether the dwell time at operating conditions for a given cell is greater than a dwell time threshold, and if it is determined that the dwell time at operating conditions for the given cell is greater than the threshold, the controller populates the given cell with data collected on-board the vehicle, and if the dwell time is not greater than the threshold, the vehicle controller communicates with an off-board network and downloads calibration data for the given cell from one or more vehicles having a dwell time at operating conditions for the given cell above the threshold.

Description

The present application relates to systems and methods for improving engine and powertrain output calibration.Engine control systems may use various calibration tables and maps to optimize engine and powertrain output as operating conditions change over a drive cycle. For example, vehicle systems may be pre-installed with engine maps used by the engine control system to determine how to schedule the various actuators. The calibration maps and tables may be populated with data collected during engine and powertrain design, testing, and experimentation. Additionally, engine control systems may be capable of adjusting and updating the calibration tables with measurements and feedback data.However, as powertrains and engines become increasingly complex, there may be many degrees of freedom in optimizing engine and powertrain output. For example, there may be various possible combinations of adjustable cam timing, adjustable intake system, adjustable valve lift, etc. In internal combustion engines with complex systems, it may take a considerable period of time (e.g. over a year of test time) to define all the possible operating combinations. Some vehicle systems may be configured for self-calibration. In this case, in-cylinder pressure sensors may be used to self calibrate the internal combustion engine from a raw (initial) internal combustion engine characteristic map. Because the combustion process is quite variable, multiple measurements must be taken at quite steady-state conditions to obtain reliable sensor output averages that can be used with confidence to update the engine map and determine control actions. However, emissions, fuel economy, and driving patterns naturally tend to be dynamic and spend little time in steady-state conditions under which adjustments could be made. As a result, with self-calibrating vehicles, many driving cycles may be required to complete the calibration.As shown by Lockwood et al. in U.S. patent application US2013 / 184966, in another example, data captured on-board a vehicle during engine operation may be processed by an on-board controller as well as by an off-board controller (such as an off-board cloud computing system). This allows for less computationally intensive on-board processing of parameters (e.g., mass adjustments) and, at the same time, more computationally intensive off-board processing of parameters (e.g., individual adjustments). The concurrent processing enables faster filling of a calibration table while maintaining processing performance and memory configuration of the on-board vehicle control system.DE 10 2013 200 260 A1 discloses systems and methods for providing more accurate torque control. DE 10 2010 018 088 A1 discloses a method for checking a calibration relating to a measuring device for detecting the surroundings of a specific vehicle. DE 10 2009 030 002 A1 discloses diagnostic control systems and in particular systems for monitoring and tracking diagnostic events.However, the inventors herein have recognized that calibration tables may not be able to be populated in a time-efficient manner even with self-calibration and off-board processing. In addition to the long periods of time for the self-calibration and off-board processing of the vehicles, there may be areas in the engine and powertrain calibration maps that remain insufficiently populated based on the driving style of the vehicle user. For example, in aggressive drivers, high speed, high load ranges of their calibration maps may be well defined, while other operating ranges are not. As another example, a driver always driving in hot and dry climates does not have enough calibration adjustments for wet and cold ambient conditions. As such, the vehicle must be adjusted within a few drive cycles before performing a fuel economy check. Furthermore, the calibration must be made fast enough for the vehicle to pass exhaust gas examinations.In one example, some of the above issues may be addressed by a method for an engine system comprising: matching data points of a vehicle powertrain calibration table using data collected on-board a vehicle and using data downloaded from an off-board network, wherein the downloaded data is collected on-board one or more other vehicles communicating with the network, determining whether the dwell time at operating conditions for a given cell is greater than a dwell threshold, and if it is determined that the dwell time at operating conditions for the given cell is greater than the threshold, the controller may populate the given cell with data collected on-board the vehicle and if the dwell time is not greater than the threshold, the vehicle controller communicates with an off-board network and downloads calibration data for the given cell from one or more vehicles having a dwell time at operating conditions for the given cell above the threshold. The off-board network may be, for example, a cloud computing system. In this way, cloud calibration may be advantageously used to more fully fill engine calibration tables in less time.For example, a first cloud calibration phase may be performed by the manufacturer during vehicle development to achieve rapid calibration to meet all emission requirements. In this case, a calibration table for a new vehicle, such as a new type (trademark or model) or a new vehicle family, can be developed. Prior to selling the vehicle to a consumer, the calibration table may be filled with calibration data collected on-board a fleet of vehicles of the same make and model developed and calibrated by the manufacturer. The calibration data collected on-board each vehicle in the fleet may be uploaded to a cloud computing system. A controller of the single vehicle may download the relevant data and rapidly update an initial calibration table of the vehicle.A second phase of cloud calibration may be performed after the vehicle is in the hands of the customer to further optimize vehicle performance characteristics for fuel economy, emissions, and drivability. The second phase of cloud calibration also accounts for aging and wear of components and allows diagnostic routines to be triggered as required. At this time, the initial calibration table (the calibration table with which the vehicle was initially delivered) may be updated. For example, a fleet of vehicles used by respective consumers may collect calibration data at various operating conditions while driving the road. The calibration data from each fleet vehicle may be uploaded to and stored in a cloud computing system. In addition, calibration data for individual vehicles may be stored in the memory of their respective controller. Each vehicle may then advantageously download data generated on-board in other vehicles having appropriate powertrain characteristics to adjust or update their respective calibration tables. For example, a first vehicle may have sufficient on-board generated data corresponding to a first portion of a powertrain calibration table of the given vehicle. Accordingly, the controller of the vehicle may populate the first area of the calibration table with the on-board collected data. Sufficient on-board data may be generated due to the first vehicle remaining for more than a time threshold in operating states (e.g., speed-load states) corresponding to the first range of the calibration table. However, the first vehicle may not have enough onboard generated data corresponding to a second, different area of the calibration table. There may not be sufficient data generated due to the vehicle experiencing less than the time threshold in operating conditions corresponding to the second region of the calibration table. As a result, the controller may determine one or more other vehicles in the fleet, such as a second vehicle having appropriate powertrain characteristics and having its calibration table sufficiently data filling the second region of the calibration table. As such, the deviation may be due to differences in driving behavior between the user of the first vehicle and that of the second vehicle. For example, the user of the first vehicle tends to perform longer highway trips, while the user of the second vehicle tends to perform shorter urban trips. Thus, while the first vehicle may have longer dwell times in high speed, high load conditions, the second vehicle may have longer dwell times in low speed, high load conditions. The controller of the first vehicle may then download the data collected on-board the second vehicle to fill the second portion of its calibration table. The vehicle actuator matches for the first vehicle may then be performed based on the updated calibration table. Thus, the calibration table of a vehicle operating on highway long-distance drives is matured with data acquired in a vehicle operating on short urban drives. As another example, the calibration tables of a vehicle operating in hot and dry weather conditions may be matched or "reif" with data captured in a vehicle operating in hot and wet weather conditions. As such, data pertaining to many different aspects of vehicle performance and adjustment may be used to improve vehicle operation.In this way, full engine maps covering a greater number of degrees of freedom may be created more quickly. By invoking data acquired on-board one or more other vehicles, calibration data regarding operating conditions and driving maneuvers that do not often occur in a given vehicle may be imported from the other vehicles. By using global data to fill the plurality of operating states in a calibration table of a vehicle, vehicle performance in these states can be improved. Additionally, local adjustments may be used to fine tune vehicle performance. As such, this allows a mean adjustment estimate to be provided more quickly while adjusting to incremental deviation more quickly by the single vehicle. By additionally using the data from one or more other vehicles during an initial phase of calibration table creation, enough data samples can be provided to provide substantially all of the speed-load points required for emissions testing. As such, this improves the level of confidence in the data populated in the vehicle's calibration table and increases the likelihood that the vehicle will pass an emissions test. Overall, engine and powertrain calibration accuracy is improved, improving vehicle performance.It should be understood that the summary above is provided to introduce in simplified form a selection of concepts that are further described in the detailed description. It is not intended to identify key or essential features of the claimed subject matter, the scope of which is defined uniquely by the claims that follow the detailed description. Furthermore, the claimed subject matter is not limited to implementations that address any of the disadvantages mentioned above or in any part of this disclosure. FIG. 1 schematically illustrates aspects of an example engine system according to an embodiment of this disclosure. FIG. 2 shows a partial view of the internal combustion engine. FIG. 3 shows a high level flow chart for adaptive calibration of a vehicle powertrain. FIG. 4 shows a flow chart for updating a calibration table according to the present disclosure. FIG. 5 shows a flow chart for adjusting the powertrain output of the vehicle based on the updated calibration table. FIG. 6 shows an exemplary calibration table adjustment.Methods and systems are provided for advancing updating one or more calibration tables used to optimize powertrain output in a vehicle system, such as the vehicle system of FIG. 1 In one non-limiting example, the engine may be configured as illustrated in FIG. 2, where the engine includes at least a cylinder, a control system, a turbocharger, and an exhaust gas recirculation system, among other features. An engine controller may be configured to execute a control routine, such as the routine of FIGS. 3-5, to update a vehicle calibration table based on on-board data and / or off-board network data and optimize vehicle powertrain output based on the updated calibration table. The off-board network data may be downloaded from an off-board network system shown in FIGS. 1-2. An example vehicle calibration table adjustment is shown in FIG. 6.FIG. 1 shows a vehicle system 100 including an internal combustion engine 10 coupled to the transmission 44. Engine 10 may be started with an engine cranking system 54 including a starter motor. The transmission 44 may be a manual transmission, an automatic transmission, or a combination thereof. The transmission 44 may include various components, such as a torque converter, a wheel drive unit, a multi-gear gear set, etc. The transmission 44 is shown coupled to the drive wheels 52 that may contact a roadway surface.In one embodiment, the vehicle system 100 may be a hybrid vehicle, wherein the transmission 44 may alternatively be powered by an electric motor 50. The motor may be, for example, a battery-powered electric motor (as shown), where the electric motor 50 is powered by energy stored in the battery 46. Other energy storage devices that may be used to power the motor 50 include a capacitor, a flywheel, a pressure vessel, etc. An energy conversion device, here inverter 48, may be configured to convert the DC output of the battery 46 to an AC output for use by the electric motor 50. The electric motor 50 may also be operated in a generator mode, i.e., as a generator, to draw energy from the vehicle motion and / or the engine and convert the drawn energy to a suitable form of energy for storage in the battery 46. Further, the electric motor 50 may be operated as a motor or generator as required to boost or receive torque during a transition of the engine 10 between different combustion modes (e.g., between spark-ignition mode and compression-ignition mode).When configured in the hybrid embodiment, the vehicle system 100 may operate in various modes, where the vehicle is propelled with the engine alone, with the motor alone, or with a combination of both. Alternatively, auxiliary or mild hybrid modes may also be employed where the engine is the primary torque source and the motor selectively adds torque during specific conditions, such as during a tip-in event. For example, engine 10 may be operated during an engine on mode and used as the primary torque source for powering wheels 52. During an engine-on mode, engine 10 may be supplied fuel from fuel system 20 including a fuel tank. The fuel tank may hold multiple fuels, such as gasoline or fuel mixtures, such as fuel with various alcohol concentrations (e.g., ethanol), including E10, E85, etc., and combinations thereof. In another example, during an engine off mode, the electric motor 50 may be operated to power the wheels. The engine off mode may be used in braking, at low speeds, during traffic lights stops, etc. In yet another example, during an "assist" mode, an alternative torque source may supplement and operate in conjunction with the torque provided by engine 10.The vehicle system 100 may further include a control system 14. The control system 14 is shown receiving information from a plurality of sensors 16 and sending control signals to a plurality of actuators 81. The control system 14 may further include a controller 12. The controller may receive input data from the various sensors or buttons, process the input data, and trigger the actuators in response to the processed input data based on instructions or code programmed therein according to one or more routines. The example control routines are described herein with reference to FIGS. 3-5.The vehicle control system may be communicatively coupled to an off-board network 13, such as a cloud computing system via wireless communication, which may be WiFi, Bluetooth, a type of cellular device, or a wireless communication protocol. As such, this connectivity to which the vehicle data is uploaded, also referred to as the "cloud", may be a commercially-used or private server on which the data is stored and which then acts optimization algorithms. The algorithms can process data from a single vehicle, fleet of vehicles, family of engines, family of powertrains, or a combination thereof. The algorithms can further account for system constraints, generate calibration data to optimize powertrain outputs, and send it back to the vehicle(s) where they are applied.The vehicle system 100 may also include an onboard navigation system 17 (e.g., a global positioning system) on the dashboard 19 with which the user may interact. The navigation system may include one or more position sensors to assist in estimating a position (e.g., geographic coordinates) of the vehicle. The dashboard 19 may further include an ignition user interface 15 via which the vehicle user may adjust the ignition status of the vehicle engine. In particular, the ignition user interface may be configured to trigger and / or end operation of the vehicle based on user input. Various embodiments of the ignition user interface may include interfaces that require a physical device, such as an active key, that must be inserted into the ignition user interface to start the engine and turn on the vehicle, or that must be removed to power down the engine and turn off the vehicle. Other embodiments may include a passive key communicatively coupled to the ignition user interface. The passive key may be configured as an electronic key fob or smart key that does not need to be plugged into or removed from the ignition user interface to operate the vehicle engine. Instead, the passive key may need to be positioned within or near the vehicle (e.g., within a distance threshold from the vehicle). Still other embodiments may additionally or optionally use a start / stop button that is manually pressed by the user to start or shut down the engine and turn the vehicle on or off. Based on the configuration of the ignition user interface, a vehicle user may provide a notification of whether the engine is in an engine on or engine off state and further whether the vehicle is in a vehicle on or vehicle off state.The controller 12 may also receive a message of the ignition status of the engine 10 from an ignition sensor (not shown) coupled to the ignition user interface. The control system 14 may be configured to send control signals to the actuators 81 based on inputs received from the sensors and the vehicle user. The various actuators may include, for example, cylinder fuel injectors, an air intake throttle coupled to the engine intake manifold, a spark plug, etc. The actuator positions may be adjusted for optimal vehicle powertrain output during engine operation based on calibration data updated using on-board data and / or off-board network data. Details of updating calibration data for optimal powertrain output are set forth in FIGS. 3-5.FIG. 2 shows an exemplary embodiment for a combustion chamber or cylinder of internal combustion engine 10 (from FIG. 1 ). The engine 10 may receive control parameters from a control system including a controller 12 and through input from a vehicle user 130 via an input device 132. In this example, the input device 132 includes an accelerator pedal and a pedal position sensor 134 for generating a proportional pedal position signal PP. As another example, input regarding a vehicle on and / or off state may be received via the ignition driver interface 15, as previously discussed with respect to FIG. 1. Cylinder (also "combustion chamber" herein) 30 of engine 10 may include combustion chamber walls 136 with piston 138 positioned therein. The piston 138 may be coupled to the crankshaft 140 such that reciprocating motion of the piston is translated into rotational motion of the crankshaft. The crankshaft 140 may be coupled to at least one wheel of the passenger vehicle via a transmission system. Further, a starter motor may be coupled to crankshaft 140 via a flywheel to enable cranking operation of engine 10.The cylinder 30 may receive intake air via a series of intake air passages 142, 144, and 146. Intake air passage 146 may communicate with other cylinders of engine 10, in addition to cylinder 30. In some embodiments, one or more of the intake ports may include a supercharger, such as a turbocharger or supercharger. For example, FIG. 2 shows engine 10 configured with a turbocharger including a compressor 174 disposed between intake ports 142 and 144 and an exhaust turbine 176 disposed along exhaust port 148. Compressor 174 may be at least partially driven by exhaust turbine 176 via a shaft 180, where the supercharger is configured as a turbocharger. However, in other examples, such as when engine 10 is provided with a supercharger, exhaust turbine 176 may optionally be omitted, where compressor 174 is driven by mechanical input from an electric motor or the engine. A throttle 20 including a throttle 64 may be provided along an intake passage of the engine to change the flow rate and / or pressure of intake air provided to the engine cylinders. For example, throttle 20 may be disposed downstream of compressor 174, or alternatively may be provided upstream of compressor 174.Exhaust passage 148 may receive exhaust gases from other cylinders of engine 10, in addition to those from cylinder 30. Exhaust gas sensor 128 is shown coupled to exhaust passage 148, upstream of emission control device 178. Sensor 128 may be selected from among various suitable sensors for providing an indication of exhaust air-fuel ratio, such as a linear oxygen sensor or a wide band oxygen sensor, a two-point oxygen sensor or oxygen sensor (EGO) (as shown), a heated oxygen sensor, a NOx, HC, or CO sensor. The emission control device 178 may be a three way catalyst (TWC), a nitrogen oxide trap, various other emission control devices, or combinations thereof.The exhaust temperature may be estimated by one or more temperature sensors (not shown) located in the exhaust passage 148. Alternatively, exhaust temperature may be inferred based on engine operating conditions, such as speed, load, air-fuel ratio (AFR), spark retard, etc. Further, exhaust temperature may be calculated by one or more exhaust gas sensors 128. It should be appreciated that the exhaust temperature may alternatively be estimated using a combination of the temperature estimation methods listed herein.Each cylinder of engine 10 may include one or more intake valves and one or more exhaust valves. For example, cylinder 30 is shown including at least one intake poppet valve 150 and at least one exhaust poppet valve 156, located in an upper portion of cylinder 30. In some embodiments, each cylinder of engine 10, including cylinder 30, may include at least two intake poppet valves and at least two exhaust poppet valves located in an upper portion of the cylinder.Intake valve 150 may be controlled by controller 12 via cam actuation via cam actuation system 151. Similarly, exhaust valve 156 may be controlled by controller 12 via cam actuation system 153. Cam actuation systems 151 and 153 may include any one or more cams and utilize one or more of cam profile switching (CPS), variable cam timing (VCT), variable valve timing (VVT), and / or variable valve lift (VVL) that may be operated by controller 12 to change valve operation. The position of intake valve 150 and exhaust valve 156 may be determined by position sensors 155 and 157, respectively. In alternative embodiments, the intake and / or exhaust valves may be controlled by electric valve actuation. For example, cylinder 30 may alternatively include an intake valve controlled via electric valve actuation and an exhaust valve controlled via cam actuation including CPS and / or VCT systems. In still other embodiments, the intake and exhaust valves may be controlled by a common valve actuator or actuation system or an adjustable valve control actuator or actuation system.The cylinder 30 may have a compression ratio that is the ratio of the volumes with the piston 138 at bottom dead center to the piston at top dead center. Conventionally, the compression ratio is in the range of 9:1 to 10:1. however, the compression ratio may be increased in some examples where other fuels are used. This can occur, for example, when fuels with a higher octane number or fuels with a higher latent enthalpy of vaporization are used. The compression ratio may also be increased if direct injection is used due to its effect on engine knock.In some embodiments, each cylinder of engine 10 may include a spark plug 192 for initiating combustion. Ignition system 190 may provide spark to combustion chamber 30 via spark plug 192 in response to spark advance signal SA (spark advance) from controller 12 under selected operating modes. However, in some embodiments, spark plug 192 may be omitted, such as where engine 10 initiates combustion by auto-ignition or by injection of fuel, as may be the case with some diesel engines.In some embodiments, each cylinder of engine 10 may be configured with one or more injectors to provide knock or pre-ignition suppressing fluid. In some embodiments, the fluid may be a fuel, wherein the injector is also referred to as a fuel injector. As a non-limiting example, cylinder 30 is shown with fuel injector 166. Fuel injector 166 is shown coupled directly to cylinder 30 to directly inject fuel therein in proportion to the pulse width of signal FPW received from controller 12 via electronic driver 168. In this manner, fuel injector 166 provides what is referred to as direct injection of fuel into combustion cylinder 30 (also referred to herein as "DI"). Although FIG. 2 shows injector 166 as a side injector, it may also be located over the piston, such as near spark plug 192. When the engine is operated with an alcohol-based fuel, such a location may improve mixing and combustion due to the lower volatility of some alcohol-based fuels. Alternatively, the injector may be located above and near the intake valve to improve mixing.Fuel may be delivered to fuel injector 166 from a high pressure fuel system 20, including fuel tanks, fuel pumps, and a fuel rail. Alternatively, fuel may be delivered from a single stage fuel pump at lower pressure, in which case the timing of direct fuel injection during the compression stroke may be more restricted than when a high pressure fuel system is used. Further, although not shown, the fuel tanks may include a pressure transducer that provides a signal to the controller 12. It should be appreciated that in an alternative embodiment, injector 166 may be a port injector that provides fuel into the intake port upstream of cylinder 30.As described above, FIG. 2 shows only one cylinder of a multi-cylinder internal combustion engine. As such, each cylinder may similarly include its own set of intake / exhaust valves, fuel injector(s), spark plug, etc.Fuel tanks in fuel system 20 may hold fuel of different qualities, such as different compositions. These differences may include different alcohol content, different octane number, different heat of vaporization, different fuel mixtures, and / or combinations thereof, etc. In one example, fuels with different alcohol contents could include a fuel that is gasoline and the other is ethanol or methanol. In another example, the engine may use gasoline as a first substance and an alcohol-containing fuel mixture such as E85 (which is about 85% ethanol and 15% gasoline) or M85 (which is about 85% methanol and 15% gasoline) as a second substance. Other alcohol containing fuels could be a mixture of alcohol and water, a mixture of alcohol, water and gasoline, etc.Controller 12 is shown in FIG. 2 as a microcomputer including microprocessor unit (CPU) 106, input / output ports 108 (I / O), an electronic storage medium for executable programs and calibration values shown as read only memory chip (ROM) 110 in this particular example, random access memory (RAM) 112, keep alive memory (KAM) 114, and a data bus. As discussed at FIG. 1, the vehicle control system that includes may be communicatively coupled to an off-board network 13, such as a cloud computing system. Controller 12 may receive various signals from sensors coupled to engine 10, in addition to those signals previously discussed, including measurement of inducted mass air flow (MAF) from mass air flow sensor 122; engine coolant temperature (ECT) from temperature sensor 116 coupled to cooling sleeve 118; a profile ignition pickup signal (PIP) from Hall effect sensor 120 (or other type) coupled to crankshaft 140; throttle position (TP) from a throttle position sensor; manifold absolute pressure signal (MAP) from sensor 124; Cylinder Air-Fuel Ratio from Oxygen Sensor 128 and Abnormal Combustion from Knock Sensor. Engine speed signal RPM may be generated by controller 12 from signal PIP. Manifold pressure signal MAP from a manifold pressure sensor may be used to provide a message of vacuum or pressure in the intake manifold. The controller may also receive user input and message regarding the engine ignition status from an ignition user interface 15.Storage medium read-only memory 110 may be programmed with computer readable data representing instructions executable by processor 106 to perform the methods described below as well as other variants that are expected but not specifically listed. The example routines are described herein with reference to FIGS. 3-5.Referring now to FIG. 3, it shows an example routine 300 illustrating a method for adaptively calibrating a vehicle powertrain. Although FIG. 3 shows an example method of updating a calibration table, the method may be applied to update one or more calibration tables that each relate to different aspects of vehicle operation.Further, it should be appreciated that the routines of FIGS. 3-4 exhibit cloud-based updating of a vehicle calibration table after the vehicle is owned and used by a consumer. As such, this may form a second phase of cloud calibration. During vehicle development, the vehicle may have undergone similar cloud-based creation and update of a calibration table at the manufacturer. As such, this may form a first phase of cloud calibration. The first calibration phase may be performed to achieve rapid calibration so that the vehicle satisfies all emissions requirements before the vehicle is sold to the consumer. During the first phase, the creation of a calibration table for a single vehicle may be advanced using calibration data collected on-board in a fleet of vehicles in use of the same make and model. Therein, as at the second stage (as embodied in FIGS. 3-4 ), on-board each vehicle of the fleet, collected calibration data may be uploaded to an off-board network, such as a cloud computing system. A controller of the vehicle undergoing the calibration updates may download the relevant data from the off-board network and fast fill (or create) an initial calibration table of the vehicle. The initial calibration table may then be further optimized when the user is using the vehicle. During this second calibration phase, the initial calibration table may be updated based not only on the user's vehicle use (i.e., onboard collected data), but also on the data collected onboard one or more other vehicles communicatively coupled to the offboard network.At 302, it may be determined whether the calibration table is scheduled for update. The calibration table may be periodically updated by a consumer during operation of the vehicle. In one example, the calibration table may be updated upon initiation of a vehicle drive cycle. In another example, the calibration table may be updated upon completion of a drive cycle of the vehicle. In yet another example, the calibration table may be updated intermittently after a time threshold has elapsed or after a distance travelled since the last calibration update during vehicle operation. In still other examples, the calibration table may be updated after expiration of a predefined number of combustion cycles since a start of the engine. Herein, the passage of combustion strokes may be counted in a vehicle controller, and when the predefined number of combustion strokes has elapsed, data collected about the predefined number of combustion strokes may be used to update the calibration table. Additionally, the data collected about the number of combustion cycles may be uploaded to an off-board network in real-time, as discussed below. In yet another example, the calibration table may be updated based on input from an on-board navigation device. As such, updating of the calibration table can be performed automatically, i.e., without consumer input. Further, updating the calibration table may be performed based on availability and connectivity to the network. It should be appreciated that one or more additional calibration tables pertaining to each different aspect of vehicle operation may be similarly updated.Upon determining that a scheduled update of the calibration table is due, at 304, the calibration table may be updated based on data collected and stored on-board the vehicle during operation of the vehicle and off-board network data. The on-board data may include data collected during vehicle operation and stored locally in the memory of the vehicle controller, such as keep alive memory.By locally storing the data, it can be quickly accessed and used to drive the updating of the vehicle's powertrain calibration table. As used herein, off-board network data may correspond to data downloaded from an off-board network. The off-board network may be a cloud computing system (also referred to herein as a data cloud) communicatively coupled to the vehicle control system. The data downloaded from the cloud computing system may have been collected on-board in one or more other vehicles communicatively coupled to the data cloud. Details of updating the calibration table will be further detailed in FIG. 4. As set forth herein, data collected on-board multiple similar vehicles enables the calibration table to be adjusted more quickly because of all of the various driving cycles in which the fleet of vehicles is operated.Once the updating of the calibration table is completed at 306, the vehicle powertrain output may be optimized based on the updated calibration table. As such, the vehicle powertrain output may include one or more of engine output, engine actuator output, transmission output, transmission actuator output, as well as battery state of charge, electric motor controls, and one or more engines installed to store and release energy in the context of a hybrid vehicle. Vehicle powertrain output optimization will be further carried out in FIG. 5. Further, at 308, data collected on-board the vehicle during the given vehicle drive cycle may be uploaded to the off-board network. In one example, data may be uploaded to the off-board network from the updated calibration table. The data may be automatically uploaded without user input (e.g., without a request from the vehicle user). For example, the data may be automatically uploaded following each drive cycle of the vehicle. Alternatively, the data may be intermittently uploaded to the data cloud in real-time, such as after a vehicle path distance threshold, a vehicle path time threshold, a threshold number of engine strokes, etc.It should be appreciated that although FIG. 3 shows data generated on-board a given vehicle being transmitted and uploaded to an off-board network (e.g., a data cloud), data generated on-board one or more other vehicles (such as in each vehicle of a fleet of vehicles on the road) may likewise be transmitted and uploaded to the off-board network. The data may then be stored in the network and shared among all vehicles communicating with the network. That is, each vehicle may have access to the data collected on-board all other vehicles of the fleet via the network.Referring to FIG. 4, it shows an example routine 400 illustrating a method for adaptive learning of the calibration table based on data collected on-board the vehicle during operation of the vehicle and data downloaded from an off-board network system, such as a cloud computing system. The calibration table may include multiple cells, and each cell may correspond to at least two operating states. Operating conditions may include at least two of engine speed, engine load, engine temperature, atmospheric pressure, fuel alcohol content, ambient humidity, ambient temperature, and hybrid battery state of charge. Still other operating conditions may include fuel octane number, fuel level, elapsed time in engine drive mode or vehicle drive mode, elapsed time with engine shut-down or vehicle shut-down, ambient temperature, train operation, fuel consumption, age of vehicle, maintenance history, etc.The output of each cell may correspond to an actuator trim or an actuator trim used to optimize vehicle powertrain output at the specific operating conditions. As one example, a cell may correspond to engine speed and engine load conditions and the output of the cell may include adjusting cylinder valve timing, cam timing, spark timing, exhaust gas recirculation rate, fuel injection amount, and / or fuel injection timing.At 402, the controller may determine a dwell time of the vehicle at operating conditions in each cell of a calibration table. The dwell time of the vehicle at a cell may be based on the duration of operation of the vehicle at the respective operating conditions over a given number of engine cycles, drive cycles, etc. Because the operating conditions may change over different drive cycles, the dwell time of the vehicle may also change at operating conditions for each cell. The deviation may be further based on vehicle user driving behavior, vehicle operation route, vehicle operation geographical location, climate conditions, etc. As vehicle operating duration in the specific operating conditions increases, dwell time for that cell may increase. For example, a vehicle operating on the highway for longer periods of time may have longer dwell times at high speed and load conditions than a vehicle operating in urban driving conditions. On the other hand, with more urban driving, the vehicle may have longer dwell times at low speed and load conditions than the vehicle traveling on the highway.In alternative examples, rather than determining a dwell time, a confidence level associated with the dwell time of a data point may be determined. Because confidence in data collected on-board a given vehicle in a given operating range is based on the dwell time of the vehicle in that operating range, there may be areas of a calibration table of the vehicle where there is low confidence in the data collected on-board. This reduces the ability of the controller to reliably optimize the powertrain output of the vehicle. In addition, if there are not enough data samples collected and averaged on-board across multiple engine speed-load points, the calibration table of the vehicle cannot be populated with sufficient confidence. On the other hand, there may be actuator control areas that can use data with low confidence levels or low dwell times if the error from the estimate has little effect on control. The inventors have recognized that using global data collected on-board one or more other vehicles, advantageously, data collected over multiple drive cycles of appropriate vehicles may be used to populate a given calibration table of the vehicle, which improves the confidence level in the data used to optimize vehicle powertrain output. In addition, there may be enough data samples with higher confidence.In determining the dwell time for each cell in the calibration table, the controller may determine 404 whether the dwell time at operating conditions for a given cell is greater than a dwell time threshold. It should be appreciated that in some examples, the dwell time may be learned as a number of combustion cycles elapsed in a given operating state, rather than as an absolute operating time at given operating states. For example, at high engine speeds, more combustion cycles take place in a shorter period of time, making the number of combustion cycles a more important aspect. Thus, in one example, it may be determined whether the vehicle has consumed more than 300 combustion cycles at the operating conditions for the given cell to statistically determine that a confidence level has been reached. The controller may determine an expected control accuracy based on the confidence range, a possible deviation from the confidence range, and an expected effect of the deviation on controlling a specific actuator, in one example. Based on the expected accuracy being sufficiently high, the controller may decide to use the network data.If it is determined that the dwell time at operating conditions for the given cell is greater than the threshold, the controller in 406 fills the given cell with data collected on-board the vehicle, in accordance with the invention. For example, the controller may match the output of the given cell with only the data collected on-board the vehicle. For example, if the output of the given cell is an adaptive gain term, the controller may adjust the value of the adaptive gain term based on the data collected on-board the vehicle. If the dwell time is not greater than the threshold, the routine may proceed to 408.In alternative examples, the controller may determine a sample size of the given cell. For example, a number of data samples or data points available for the given cell may be determined. If the number of samples is greater than a sample size threshold, the controller may adjust the value of the adaptive gain term based on the data collected on-board the vehicle. Otherwise, the routine may proceed to 408.At 408, the vehicle controller interfaces with an off-board network, such as a cloud computing system, and download calibration data for the given cell from one or more vehicles having a dwell time at operating conditions for the given cell above the threshold. As such, the data may be downloaded from one or more (other) vehicles having powertrain configurations matching those of the given vehicle. For example, data may be present that is generated on-board one or more other vehicles with appropriate powertrain configuration that is relayed to and stored on the off-board network. The controller may determine one or more vehicles with appropriate powertrain configurations and further determine whether, in any of these vehicles, the corresponding cell(s) of their calibration table is / are populated with data collected during operation with a dwell time above the threshold in the corresponding operating range. The controller may also notice the dwell time for each of these vehicles at the operating conditions corresponding to the given cell. It should be appreciated that in alternative embodiments, the data processing and computations may be performed on the off-board network rather than on the vehicle controller level. The greater computing power of the off-board network and cloud computing system would allow for a faster build of calibration tables for vehicles with similar operating conditions. Additionally, issues associated with peer-to-peer communication may be reduced, such as calibration in one of the vehicles drifting out in a wrong direction due to mechanical wear or degradation.For example, the off-board network may track all weather conditions (such as humidity), and the vehicle controller may recover from such influence, apply current weather for the day, and adjust data to be downloaded back to the vehicle based on how the tracked weather parameter affects operation of the vehicle powertrain. As another example, the off-board network may track local traffic conditions. The vehicle controller may then adapt vehicle operation to the local traffic conditions. For example, the controller may adjust a pedal response curve based on heavy traffic volume.Next, at 410, calibration data for the given cell may be downloaded from the off-board network. In particular, data is downloaded from a cloud computing system for the given cell that is collected on-board one or more other vehicles that have longer than a dwell time threshold at operating conditions for the given cell.In alternative examples, the confidence level or dwell time for each cell may be compared to a confidence level threshold. If the confidence level threshold is exceeded, the on-board collected data may be considered reliable. Additionally, for some portions of the calibration table or for some actuator control portions, as discussed above, data may also be used that has a dwell time or confidence level below the confidence level threshold if the error from the data collected in that cell has little to no effect on controlling the given actuator.Next, at 412, the calibration table for the given vehicle may be adjusted based on the downloaded data. In one example, an average of the downloaded data collected on-board one or more of the determined vehicles may be used to match the calibration table. In another example, the off-board calibration data may be assigned a weight based on a dwell time for each determined vehicle at the operating conditions for the given cell. The weighted data from all determined vehicles may be processed to determine values for the given cell in the calibration table. The assigned weight may be based on the absolute dwell time for each vehicle at the operating conditions for the given cell. Alternatively, the assigned weight may be based on the dwell time relative to the threshold or the dwell time for each determined vehicle relative to the dwell time of the given vehicle (which has less than the dwell time threshold). For example, if the dwell time for a vehicle determined in the off-board network from a fleet of vehicles exceeds the dwell time of the vehicle traversing the calibration updates, the weight of the data collected on-board the determined vehicle may be increased. As such, the off-board network may track individual vehicles as well as the quality of data captured from each vehicle. Furthermore, a course of calibration changes can also be monitored and outliers can be determined for calibrations with which the values are to be returned outside a threshold value determined by the network. The outliers cannot be used to update the calibration table. In addition to the dwell time, the weighting factor may also be based on the number of available data samples. A confidence value may be calculated based on the dwell time and the number of available data samples, and the weighting factor may be adjusted based on the confidence value. Thus, data samples downloaded from the off-board network having a higher confidence value may be weighted more heavily, while those having a lower confidence value may be weighted more lightly.As will be discussed below, the weighting factor may be further based on one or more vehicle conditions, such as a location of the vehicle. For example, the weighting factor of data samples collected on-board one or more vehicles that are closer (e.g., within a distance threshold) to the given vehicle undergoing an update to the calibration table may be stronger, while the weighting factor of data samples collected in a vehicle that is farther (e.g., farther than a distance threshold from the given vehicle) may be weaker. As discussed above, the data processing and computations may be performed in the off-board network instead of at the vehicle controller level. In one example, performing the computations in the cloud computing system may reduce the risk that a misrouted vehicle will affect its neighbors with bad data. For example, the calibrations may be normalized for local climate conditions. Weather, altitude, humidity, etc. may be tracked in the cloud, the effects of the weather related parameters on vehicle performance backed off, and local effects applied and downloaded. Thus, the cloud computing system may substantially determine which data points collected on-board the fleet of vehicles are reliable for downloading and updating the calibration table.Subsequently, the output of the given cell (which has on-board data with dwell time below the threshold) may be matched with the downloaded off-board network data. For example, the cell may be populated with data based on data collected on-board one or more other vehicles having dwell times above the threshold for the given cell. In an alternative example, the output of the given cell may be adjusted based on a combination of vehicle-learned data and the downloaded off-board network data. In this example, the on-board and off-board network data may be given a weight based on the dwell time of the respective vehicle in the operating states for the given cell. The weighted on-board vehicle data and the off-board network data may be processed to determine an output (e.g., an adaptive expression) for the given cell. Here, the on-board generated data may be given a weaker weight, while the off-board network data may be given a stronger weight, where the weaker weight is weakened as the dwell time of the given vehicle is reduced, and the stronger weight is strengthened as the dwell time of the corresponding vehicle is increased. Thus, the contribution of each vehicle's data to calibration table update may be scaled based on the dwell time of the vehicle in specific operating conditions. In this way, values in each cell of the calibration table may be updated based on adaptive learning using on-board vehicle data, off-board network data, or both. If the vehicle dwell time is below the threshold for a given cell, the updated value may be matched with data obtained from the off-board network or with a combination of on-board vehicle data and off-board network data. If the dwell time of the vehicle is above the threshold for a given cell, the updated value may be obtained based only on onboard vehicle data. As such, the calibration tables may be normalized for local climate conditions. For example, climate conditions such as weather, altitude, humidity, etc. may be tracked for each vehicle. Effects of these deviations in the downloaded data may be nullified and local climate conditions for the given vehicle may be applied. Steps 402- 412 may be repeated for each cell of the vehicle's calibration table until the entire table has been sufficiently updated. After updating each cell of the calibration table, the routine may return to 306 in FIG. 3. The routine of FIG. 3 may likewise be repeated for multiple calibration tables, each controlling different aspects of vehicle operation.As an example, data collected on-board a trading party's vehicle that is long distances and frequently on the highway may be used to make the calibration table of another vehicle "reif" that is operated for daily utility trips on short urban trips. Thus, the global data may be used to fill a plurality of operating states and corresponding cells. Furthermore, local adjustments could be made to fine tune a particular vehicle and make it more robust to vehicle-to-vehicle variations. For example, the controller may communicate with the off-board network to determine a quantity of vehicles in a fleet having a longer dwell time at selected operating conditions. The controller may then select a subset of the vehicle from the set based on the geographic location relative to a location of the given vehicle (at which the calibration table is updated). The controller may then update the calibration table based only on the data collected on-board the vehicle subset. Alternatively, the controller may update the calibration table using data collected on-board all determined vehicles, but weighting the data, giving the data collected on-board the subset of vehicles a higher weight. In alternative examples, the selecting may be performed by the off-board network instead of at the controller level. As such, the cloud computing system may have the largest amount of data and greater computing power. The cloud computing system may filter individual inputs by quality and determine which amount of data is most suitable and reliable for a particular vehicle, while setting the weight for each amount of data. After the amount of data is selected for download by the cloud computing system, the vehicle controller may download the selected data and update the calibration table of the given vehicle.In still further examples, the vehicle data cloud may store data pertaining to many additional aspects of vehicle performance and adjustment. As one example, data for replenishing fuel may be stored and shared. Herein, the controller may transmit data to the off-board network regarding vehicle operating state at a time of refueling, location of refueling, a particular refueling location where refueling occurred, as well as details of fuel replenished to the fuel tank (e.g., fuel alcohol content, fuel octane number, etc.). The stored data may then be retrieved from other fleet vehicles when refueling at the same gas station to balance vehicle operation. For example, powertrain actuator settings for another vehicle refilling at the same gas station may be adjusted based on the stored data to better address transitions (e.g., torque transitions) resulting from fuel differences between the fuel available at the gas station and the fuel currently in the vehicle's fuel tank.It should be appreciated that although the calibration table is updated before vehicle optimization is triggered in some examples, in alternative examples, the vehicle may be started with a calibration table that is mature enough to allow driving and optimization in principle. The table is then rapidly updated with data downloaded from the data cloud. As such, this reduces the calibration effort of the vehicle controller. For example, the vehicle may be started and initially operated for a number of engine cycles with a base calibration table based only on onboard collected data, and then the calibration table may be updated with data downloaded from the offboard network. During the creation of the initial calibration table, during a first calibration phase, a base calibration table may be populated with data collected on-board the vehicle during vehicle testing, the data collected over a number of engine cycles, and may be further populated with data downloaded from the off-board network, the downloaded data collected on-board a number of vehicles of the same or similar make and model developed by the manufacturer (e.g., proprietary vehicles of the same make and model in the vehicle fleet of the manufacturer, which are driven around by engineers during a development phase of the vehicle). In one example, the vehicle may adjust the calibration table within 2 emissions test drive cycles so that the vehicle may pass through a third emissions test drive cycle and meet the emissions standard. This allows the vehicle to adapt to emissions standards rapidly during the development phase at the manufacturer and before the vehicle is sold to the consumer.Further, although the routine of FIG. 4 shows updating cells of a calibration table of the vehicle with off-board network data when the vehicle has a short dwell time in those cells, it should be appreciated that cells in which the vehicle has a sufficient dwell time may also be updated based on the off-board network data in yet other examples. For example, the vehicle controller may compare the dwell time and an associated confidence value for data collected on-board the vehicle for a given cell with the dwell time and the associated confidence value for data collected on-board one or more other vehicles for the given cell. The one or more vehicles may be vehicles having suitable characteristics and may be further located at a similar location to the given vehicle (e.g., within a distance threshold). If multiple vehicles and data samples are available for the given cell, the vehicle controller may calculate how far the on-board data is from the data sample average of the multiple vehicles to accurately distinguish whether the given vehicle is operating far from the data average. If so, the vehicle controller may activate specific diagnostic routines.In an example illustration, a vehicle controller may trigger one or more diagnostics for vehicle propulsion components based on a comparison of data points of a cell of a vehicle propulsion calibration table collected on-board a vehicle with data points for the cell downloaded from an off-board network, where the downloaded data is collected on-board one or more other vehicles communicating with the network. The one or more other vehicles may have appropriate powertrain characteristics and be within a distance threshold of the given vehicle. The controller may compare an on-board collected data point of the cell to a statistically significant mean (e.g., mean, modal, median, weighted mean, etc.) of the downloaded data. In response to a difference between the cell data point and the statistically significant mean being greater than a threshold, a diagnostic routine may be triggered.In another example illustration, a vehicle controller may match data points of a vehicle drive calibration table using data collected on-board a vehicle and using a selected amount of data downloaded from an off-board network, where the downloaded data is collected on-board one or more other vehicles communicating with the network. The selected amount of data may be selected by a control module of the off-board network based on various factors.In yet another exemplary illustration, during a first calibration phase, data points of a vehicle drive calibration table may be adjusted using data collected on-board the vehicle during a vehicle test and further adjusted using data downloaded from an off-board network, the downloaded data collected on-board one or more other vehicles (e.g., fleet) communicating with the network, the one or more vehicles having appropriate drive configurations, the downloaded data collected while the one or more vehicles are being developed by the manufacturer. The first calibration phase may be performed while the vehicle is at the manufacturer's location and before the vehicle is used by a consumer. A first, initial (or base) calibration table may be created at the end of the first calibration phase. Additionally, data points of the first calibration table may be adjusted so that the vehicle meets emissions standards. Then, during a second calibration phase, data points of the vehicle propulsion calibration table may be adjusted using vehicle on-board data collected during vehicle use by a consumer and further adjusted using data downloaded from an off-board network, the downloaded data collected on-board one or more other vehicles (e.g., fleet) communicating with the network, the one or more vehicles having appropriate propulsion configurations, the downloaded data collected while the one or more vehicles are being used by respective consumers. The second calibration phase may be performed after the vehicle has left the manufacturer location and while the vehicle is being used by the consumer. A second, updated calibration table may be created at the end of the second calibration phase. Additionally, data points of the second calibration table may be adjusted to optimize the vehicle for use by the user. During the first and second calibration phases, data may be selected for download and use in calibration table matching in an off-board network. The selecting may be based on various factors, such as the quality of individual inputs, the location of data collection, the surrounding weather conditions at the time of data collection, the history of data collection in each vehicle of the fleet, etc. The off-board network may set a weighting factor for each amount of data collected on-board a vehicle of the fleet. The off-board network may then select one or more sets of data for download to the calibration table of the particular vehicle, the selection being performed based on what is most suitable and reliable for the particular vehicle. In this way, a vehicle controller may use calibration information from an off-board network to improve or verify the given calibration table of the vehicle during operating conditions where the vehicle does not have enough on-board data. By utilizing data downloaded from an off-board network and collecting the on-board of one or more vehicles communicating with the off-board network, the calibration table can be updated more quickly and reliably. Overall, the vehicle may adapt more quickly to deviations in global and regional operating states.FIG. 5 shows an example routine 500 illustrating a method for adjusting a vehicle powertrain parameter, such as actuator setting, for optimizing vehicle powertrain output based on adaptive learning of the vehicle calibration table.At 502, the controller may estimate and / or calculate engine operating conditions. These may include, for example:Next, at 504, the controller may determine actuator settings for optimal vehicle powertrain output at the estimated operating states based on the updated calibration table. For example, the controller may determine the output corresponding to the estimated operating conditions of one or more cells of the updated calibration table. As such, the output of the cells may include vehicle powertrain output. The vehicle powertrain output may include one or more engine outputs, transmission outputs, hybrid electric motor outputs, and outputs of other engines used to store and release energy. For example, the engine output may include one or more of the following: scheduled engine speed and engine load; while the transmission output may include one or more of the following: scheduled gear selection, transmission shift scheduling, transmission pressure control, and torque converter control. Hybrid electric motor output may include torque command output for the electric motor. The output of the cells as read by the controller may include absolute actuator settings or adjustment terms, such as an adaptive gain term for actuator setting.Next, at 506, the controller may adjust one or more vehicle powertrain actuators based on the determined settings. This includes matching one or more engine actuator outputs, transmission actuator outputs, and hybrid electric motor outputs to optimize vehicle powertrain output. The adjusted settings of engine actuator outputs may include one or more of: valve timing, cam timing, injection timing, injection amount, spark timing, exhaust gas recirculation rate, and boost pressure. Similarly, adjusted settings of transmission actuator outputs may include one or more of transmission gear selection, transmission clutch pressure, torque converter clutch pressure, electric oil pump pressure, line pressures and solenoid response times, transmission line pressure, shift solenoid characteristics, torque converter lock-up or slip rates, etc. For hybrid electric motor output, the motor controller torque command may count.In this way, global calibration data collected in one or more vehicles (e.g., a fleet of vehicles having appropriate configurations) during vehicle use by respective consumers may be utilized to optimize the vehicle powertrain output of a given vehicle used by its respective consumer. By uploading collected data on-board of each vehicle to an off-board network, the collected data may be shared by each vehicle of the fleet communicating with the off-board network. This allows the calibration table of each vehicle to be verified with data collected on-board the given vehicle as well as data collected during operation of other vehicles. In addition to storing the calibration data and corresponding dwell times, the off-board network may store additional vehicle performance and adjustment data. For example, the off-board network may be used to store vehicle performance and adjustment information after refueling at a particular gas station. Another vehicle refueling at the same gas station could use the previous vehicle information to adapt to fuel differences from current conditions in the tank.In one example, a method for adaptively calibrating vehicle powertrain output may include: uploading to an off-board network data collected on-board a first and second vehicle with appropriate powertrain characteristics; populating a first portion of a first calibration table of the first vehicle using data collected on-board the first vehicle; and downloading data collected on-board the second vehicle from the off-board network to populate a second portion of the first calibration table. The first range of the first calibration table corresponds to a first set of operating conditions, and filling the first range using data collected on-board the first vehicle occurs in response to the first vehicle being in the first set of operating conditions for greater than a threshold time. The second range of the first calibration table corresponds to a second set of operating conditions that is different than the first set of operating conditions, and filling the second range using data collected on-board the second vehicle occurs in response to the first vehicle being less than the threshold time in the second set of operating conditions and the second vehicle being more than the threshold time in the second set of operating conditions. The method may further include: uploading, to the off-board network, data collected on-board a third vehicle having appropriate powertrain characteristics; and further filling the second region of the first calibration table of the first vehicle using weighted on-board the data collected in the third vehicle, the weighted data based on a dwell time of the third vehicle in the second operating state relative to both the first and second vehicles. Further, during operation of the first vehicle, the settings for one or more powertrain actuators may be adjusted based on the first calibration table populated with data collected on-board the first and second vehicles.Furthermore, the data of the second and third vehicles could be averaged together. Still further, it may be possible to revert the effects of data collected on-board the second and third vehicles to different powertrain combinations and then apply the balanced data to different vehicle powertrains. For example, gasoline obtained at a gas station from a vehicle with a V8 internal combustion engine may have an effect on the operation of the internal combustion engine, such as evaporation. This effect can be learned and understood. If it is being fueled at the same gas station, this "learned effect" could then be applied to a vehicle with an I4 internal combustion engine because it is known that it will affect it in a similar known manner.In yet another representation, a method of adjusting a vehicle calibration table includes matching data points of a vehicle powertrain calibration table using data collected on-board a vehicle over a number of drive cycles and using data downloaded from an off-board network, the downloaded data collected on-board one or more other vehicles communicating with the network over a number of drive cycles, vehicle operating conditions covered by the vehicle over the number of drive cycles partially overlapping or non-overlapping with vehicle operating conditions covered by the one or more other vehicles. The one or more vehicles may be selected based on a dwell time in the vehicle operating conditions. The one or more vehicles may be further selected or weighted based on an operating location relative to the location of operation of the given vehicle.As another example, a method for updating a vehicle powertrain calibration table includes: uploading to an off-board network data collected on-board of each of the first, second, and third vehicles with appropriate powertrain characteristics; filling a first portion of a first calibration table of the first vehicle using data collected on-board of the first vehicle; and downloading data collected on-board of the second vehicle from the off-board network to fill a second portion of the first calibration table during a first state while downloading data collected on-board of the third vehicle from the off-board network to fill the second portion of the first calibration table during a second state. During the first state, a dwell time of the second vehicle in the second region of the first calibration table may be longer, while in the second state, a dwell time of the third vehicle in the second region of the first calibration table may be longer.Alternatively, the dwell time of both the second and third vehicles may be longer, and during the first state, the second vehicle may be selected based on a geographic location of the second vehicle relative to the first vehicle (e.g., the second vehicle is closer to or less than a distance threshold from the first vehicle). In comparison, during the second state, the third vehicle may be selected based on a geographic location of the third vehicle relative to the first vehicle (e.g., the second vehicle is closer to or less than a distance threshold from the first vehicle). Still further, the dwell time of both the second and third vehicles may be longer, and during the first state, data collected on-board the second vehicle may be weighted more heavily than data collected on-board the third vehicle due to the geographic location of the second vehicle relative to the first vehicle (compared to the relative location of the third vehicle). In comparison, during the second state, data collected on-board the third vehicle may be weighted more heavily than data collected on-board the second vehicle due to the geographic location of the third vehicle relative to the first vehicle (compared to the relative location of the second vehicle). Alternatively, the data processing may be performed at the off-board network level instead of at the vehicle controller level. In this case, the higher processing capability of the cloud computing system can advantageously be used to perform weighted averaging of data collected on-board many vehicles in order to find an optimum (e.g. "best") calibration.In yet another example, the third vehicle controller may upload collected data to the off-board network with appropriate powertrain characteristics; and further fill the first portion of the first calibration table of the first vehicle using weighted data collected on-board the third vehicle, the weighted data based on a dwell time of the third vehicle in the second operating state relative to the first vehicle. The weighted data may be further based on a location of the third vehicle relative to the first vehicle. For example, data collected on-board the first vehicle may be weighted more heavily as a difference between the data collected on-board the first vehicle and the data collected on-board the third vehicle is less. Otherwise, as the difference is greater, the data collected on-board the third vehicle may be weighted more heavily. Additionally, a powertrain component diagnostic routine may be initiated in the first vehicle in response to the difference being greater.The steps in FIGS. 3-4 are further performed by an example calibration table update indicated in map 600 in FIG. 6. Specifically, the map 600 shows an example calibration table 602 of the vehicle "a" before being updated by the vehicle controller, an example amount of data 604 from the off-board network, and an updated calibration table 606 of the vehicle "a". The amount of data from the off-board network may include amounts of data for vehicles "b", "c", and "d", which may have powertrain characteristics matching vehicle "a". Each calibration table may include a plurality of cells, each of the plurality of cells may correspond to at least two operating conditions, the operating conditions including at least two of engine speed, engine load, engine temperature, atmospheric pressure, fuel alcohol content, ambient humidity, and others. The calibration table data set shown here is a two-dimensional map recorded as a function of engine speed and load. The amount of data from the off-board network (including data for vehicles "b", "c", and "d") corresponds to the speed-load range represented in the calibration table of vehicle "a". In some examples, the calibration table may be recorded in a three-dimensional map as a function of at least three operating conditions, the operating conditions including at least three of engine speed, engine load, engine temperature, atmospheric pressure, fuel alcohol content, ambient humidity, and others.The calibration table may include on-board data points, such as a 1- a 8 for vehicle "a", b 1- b 8 for vehicle "b", c 1- c 8 for vehicle "c", d 1- d 8 for vehicle "d", and e 1- e 8 for vehicle "e". As such, data points a 1- a 8 correspond to data points collected on-board of vehicle "a" over multiple vehicle drive cycles, data points b 1- b 8 correspond to data points collected on-board of vehicle "b" over multiple vehicle drive cycles, and so forth for vehicles "c", "d", and "e". Each cell in the calibration table may include at least one data point. The data point may include an absolute actuator setting, an adaptive expression, etc. Data sets for vehicles b, c, d, and e are present in the off-board network. The data amount for the vehicle "a" is an on-board data amount collected when the vehicle "a" was in operation at a consumer. Each cell may be associated with a set of operating conditions, and the controller may further have information regarding the dwell time of the given vehicle at the operating conditions for the given cell. As such, the dwell time corresponds to the amount of time the vehicle has elapsed over one or more vehicle drive cycles in the operating states corresponding to the given cell. Data points collected in the vehicle at operating conditions of the cell shorter than a dwell threshold are shown in boxes outlined with dashed lines. These data points may be less mature and require further updating to make them more reliable. In comparison, data points collected in the vehicle at operating conditions of the cell greater than a dwell threshold are shown in dashed-line boxes. These data points may be more mature and may require less updating (or no further updating) to make them more reliable.Prior to updating, the calibration table 602 of the vehicle "a" may include the data points a 1- a 8. The data points a1 - a4 represent data collected in the speed-load range in which the vehicle "a" has been operated for a duration longer than the threshold. Thus, data points a1 - a4 were collected in the vehicle with longer dwell times (longer than the threshold). The data points a 5-a 8 represent data collected in the speed-load range in which the vehicle "a" has been operated for a duration shorter than the threshold. Thus, data points a 5-a 8 may have been collected in the vehicle with shorter dwell times.In the off-board network, vehicles "b", "c", "d", and "e" may have been operated with dwell times longer than the threshold at operating states of cells corresponding to data points b 1- b 2, b 5- b 8, c 1- c 4, c 7- c 8, d 1- d 4, d 7- d 8, and e 1- e 6, respectively. That is, in the speed-load range in which the vehicle "a" has data points with dwell times shorter than the threshold, at least the vehicles "b", "c", and "d" have dwell times longer than the threshold. In other words: In the speed-load range in which the vehicle "a" has not been operated for a duration above the threshold value, at least one of the vehicles "b", "c" and "d" has been operated for a duration above the threshold value and thus has adapted calibration data for these ranges.Because the vehicle "a" has dwell times at operating states of the cells having the data points a 1- a 4 greater than the threshold value, during the updating of the calibration table, the calibration table of the vehicle "a" can be filled with on-board data points a 1- a 4 in these cells. However, the vehicle "a" has dwell times shorter than the threshold value in the operating states of the cells having the data points a5-a8. Thus, the value of the cells may be matched with data points a5-a8 based on data from vehicles "b", "c", and "d" downloaded from the off-board network. For example, the matched data point Δa5may be a function of b5and e5because the vehicles "b" and "e" have dwell times longer than the threshold at operating states of the cells having data points b5and e5. Similarly, the adjusted data point Δa6may be a function of b6and e6, the adjusted data point Δa7may be a function of b7, c7and d7, and the adjusted data point Δa8may be a function of b8, c8and d8. In one example, the vehicle "a" calibration table may be adjusted based on on-board data and off-board network data weighted according to their respective dwell times. In this example, Δa5may be a function of a5, b5, and e5, Δa6may be a function of a6, b6, and e6, Δa7may be a function of a7, b7, d7, and e7, and Δa8may be a function of a8, b8, c8, and d8.In one example, the calibration table may be adjusted based on an optimized set (e.g., statistical mean) of data points at operating conditions of a given cell downloaded from the off-board network. In another example, a confidence value may be calculated based on the number of data points available for operating states of a given cell in the off-board network. Further, a weight may be assigned to the data point based on the confidence value, and a weighted average of all data points at operating conditions of a given cell may be calculated. In yet another example, the calibration table may be updated based on the confidence value for on-board collected data. In this example, a data point for a given cell may be assigned a confidence value based on operation of the vehicle at operating conditions of the given cell. The confidence value of the on-board data point may be compared to a confidence value of each data point in the off-board network at operating conditions of the given cell to determine the difference between the confidence value of the on-board data and the confidence value of the off-board data. If the difference is greater than a threshold, the higher confidence data (on-board data or off-board data) may be used to update the calibration table. If the difference is less than a threshold, an average of on-board data and off-board data may be calculated to update the calibration table.In this way, data collected on-board multiple vehicles and shared among them over a network may be advantageously used to propel updating of a vehicle's calibration table. In particular, a large plurality of operating states can be covered in a shorter period of time because of all the different driving cycles in which the fleet of vehicles is operated. By utilizing off-board network data to optimize the on-board calibration table for a particular vehicle, the calibration table for the vehicle (operating in a narrow range of operating conditions) can be made to operate in a wider range of operating conditions. By using weighted global data to match individual cells of a powertrain calibration table of a given vehicle, an average adaptation time may be reduced while improving powertrain calibration accuracy.It should be noted that the example control and estimation routines included herein may be used with various engine and / or vehicle system configurations. The control methods and routines disclosed herein may be stored as executable instructions in non-transitory memory. The specific routines described herein may represent one or more of any number of processing strategies such as event-driven, interrupt-driven, multi-tasking, multi-threading strategies, and the like. As such, various illustrated actions, operations, and / or functions may be performed in the order illustrated, in parallel, or in some cases omitted. Likewise, the order of processing is not necessarily required to achieve the features and advantages of the embodiments described herein, but is provided for ease of illustration and description. One or more of the illustrated actions, operations, and / or functions may be repeatedly performed depending on the particular strategy being used. Further, the described actions, operations, and / or functions may graphically represent code to be programmed into the non-transitory memory of the computer readable storage medium in the engine control system.It should be understood that the configurations and routines disclosed herein are exemplary in nature and that these specific embodiments are not to be considered in a limiting sense because numerous variations are possible. For example, the above technology can be applied to six-cylinder V engines (V-6), inline four-cylinder engines (I-4), inline six-cylinder engines (I-6), twelve-cylinder V engines (V-12), horizontally opposed four-cylinder engines (Opposed 4), and other types of internal combustion engines. The subject matter of the present disclosure includes all novel and non-obvious combinations and sub-combinations of the various systems and configurations, and other features, functions, and / or characteristics disclosed herein.The following claims particularly point out certain combinations and sub-combinations which are considered novel and not obvious. These claims may refer to "a" element or "a first" element, or equivalents thereof. Such claims should be understood to include integration of one or more such elements, neither requiring nor excluding two or more such elements. Other combinations and sub-combinations of the disclosed features, functions, elements, and / or characteristics may be claimed by the following of the present claims or by the presentation of new claims in this or a related application. Such claims, whether broader, narrower or different from the scope of the original claims, are also considered to be included within the subject matter of the present disclosure.

Claims

A method comprising: matching data points of a vehicle powertrain calibration table using data collected on-board a vehicle (100) and using data downloaded from an off-board network (13), wherein the downloaded data is collected on-board one or more other vehicles (100) communicating with the network (13), wherein it is determined whether the dwell time at operating conditions for a given cell is greater than a dwell threshold, and if it is determined that the dwell time at operating conditions for the given cell is greater than the threshold, the controller fills the given cell with data collected on-board the vehicle, and if the dwell time is not greater than the threshold, the vehicle controller communicates with an off-board network and downloads calibration data for the given cell from one or more vehicles having a dwell time at operating conditions for the given cell above the threshold.The method of claim 1, wherein the matching data points includes matching each data point of the vehicle powertrain calibration table, and wherein each data point of the powertrain calibration table includes either an engine output or a transmission output (44).The method of claim 1, wherein the data collected on-board the vehicle (100) includes data collected on-board the vehicle (100) used by a consumer, and wherein the downloaded data is collected on-board one or more other vehicles (100) used by respective consumers.The method of claim 1, wherein the calibration table includes a plurality of cells, each of the plurality of cells corresponding to at least two operating conditions, the operating conditions including at least two of engine speed, engine load, engine temperature, atmospheric pressure, fuel alcohol content, ambient humidity, and fuel octane number.The method of claim 4, wherein matching data points of the calibration table includes filling the calibration table with data collected on-board the vehicle (100) at operating conditions for the given cell when a dwell time of the vehicle (100) at operating conditions for a given cell is greater than a threshold.The method of claim 5, wherein the matching data points of the calibration table further includes, if a dwell time of the vehicle (100) at operating conditions for a given cell is less than the threshold, downloading, for the given cell, data collected on-board one or more other vehicles (100) having a dwell time greater than the threshold at the operating conditions for the given cell, and filling the given cell of the calibration table based on the downloaded data.The method of claim 1, wherein the off-board network (13) includes a cloud computing system.The method of claim 1, wherein the vehicle (100) also communicates with the off-board network (13), further comprising uploading data collected on-board the vehicle (100) to the off-board network (13).The method of claim 1, wherein an output of the vehicle powertrain calibration table includes one of engine output, engine actuator output, transmission output (44), transmission actuator output, and hybrid electric motor output (50).The method of claim 9, wherein the engine output includes one or more of: engine speed and engine load; wherein the engine actuator output includes one or more of: valve timing, cam control, injection control, injection amount, exhaust gas recirculation rate, fuel injection split ratio, spark advance, and boost pressure; wherein the transmission output (44) includes one or more of: gear selection and shift schedule; and wherein the transmission actuator output includes one or more of: transmission line pressure, shift solenoid characteristics, torque converter lock-up rate, transmission clutch slip rate.The method of claim 1, wherein the one or more vehicles (100) have powertrain characteristics matching the powertrain characteristics of the vehicle (100), and wherein the matching includes automatically matching during initiation of a vehicle drive cycle, termination of a vehicle drive cycle, or during a scheduled update.The method of claim 1, further comprising during vehicle operation, adjusting one or more engine and powertrain actuators based on engine operating conditions and further based on the adjusted vehicle calibration table.A method comprising: uploading data collected on-board both the first and second vehicles (100) with appropriate powertrain characteristics to an off-board network (13); filling a first portion of a first calibration table of the first vehicle (100) using the data collected on-board the first vehicle (100); downloading the data collected on-board the second vehicle (100) from the off-board network (13) to fill a second portion of the first calibration table; Wherein it is determined whether the dwell time at operating conditions for a given cell is greater than a dwell threshold and if it is determined that the dwell time at operating conditions for the given cell is greater than the threshold, the controller fills the given cell with data collected on-board the vehicle, and if the dwell time is not greater than the threshold, the vehicle controller communicates with an off-board network and downloads calibration data for the given cell from one or more vehicles having a dwell time at operating conditions for the given cell above the threshold.The method of claim 13, wherein the first range of the first calibration table corresponds to a first set of operating conditions, and wherein filling the first range using data collected on-board the first vehicle (100) occurs in response to the first vehicle (100) being in the first set of operating conditions for longer than a threshold time.The method of claim 14, wherein the second range of the first calibration table corresponds to a second set of operating conditions that is different from the first set of operating conditions, and wherein filling the second range using data collected on-board the second vehicle (100) occurs in response to the first vehicle (100) being within the second set of operating conditions less than the threshold time and the second vehicle (100) being within the second set of operating conditions greater than the threshold time.The method of claim 15, further comprising: uploading data collected on-board a third vehicle (100) having appropriate powertrain characteristics to the off-board network (13); and further filling the second region of the first calibration table of the first vehicle (100) using weighted data collected on-board the third vehicle (100), the weighted data based on a dwell time of the third vehicle (100) in the second set of operating conditions relative to both the first and second vehicles (100).The method of claim 16, wherein the calibration table is a vehicle powertrain calibration table, the method further comprising, during operation of the first vehicle (100), adjusting settings for one or more powertrain actuators (81) based on the first calibration table populated with data collected on-board of both the first and second vehicles (100).A vehicle system comprising: an engine (10); a powertrain (12) coupled between the engine (10) and vehicle wheels (52); one or more actuators (81) configured to change powertrain output; a communication module to communicatively couple the vehicle system (100) to an off-board cloud network (13); and a controller (12) having computer readable instructions embodied in non-transitory memory for: storing on-board data of the vehicle (100) locally generated while also uploading the on-board generated data to the off-board cloud network (13); based on operating conditions at which the data has been generated on-board the vehicle (100) and further based on a dwell time of the vehicle (100) at the operating conditions, filling one or more cells of a powertrain calibration table; and downloading data corresponding to unfilled cells of the calibration table from the off-board cloud network (13), wherein the downloaded data has been generated on-board one or more other vehicles (100); and adjusting settings for the one or more actuators (81) based on the calibration table.The system of claim 18, wherein filling based on operating conditions and a dwell time of the vehicle (100) includes filling a cell with the on-board generated data if a dwell time of the vehicle (100) at an operating condition corresponding to the cell is greater than a threshold, and not filling the cell with the on-board generated data if the dwell time of the vehicle (100) at an operating condition corresponding to the cell is less than the threshold.The system of claim 19, wherein the downloading includes downloading data generated on-board one or more other vehicles (100) having a longer dwell time at operating conditions than the threshold corresponding to unfilled cells of the calibration table, the one or more other vehicles (100) having powertrain characteristics that match those of the vehicle (100).

Images