METHOD FOR MONITORING COMPONENT LIFESPAN

Abstract

Procedures for a vehicle, including: Recursive prediction of a deterioration state of a vehicle component by updating a previously detected deterioration state of the vehicle component, wherein the update is based on a vehicle operating parameter detected by a sensor, wherein the previously detected deterioration state is based on a specific metric, including a previous history of the specific metric; and Converting the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator on a display, wherein the conversion is based on previous driving history data and on predicted future driving, including the previous history of the specified metric; wherein the vehicle component is an intake air filter of the internal combustion engine and the determined metric is one or more of a mean and a standard deviation value of airflow through the filter and the detected vehicle operating parameter includes manifold airflow.

Description

Area

[0001] The present application relates to methods carried out in vehicles, such as hybrid vehicles, to estimate the remaining service life of a vehicle component using statistical predictions. Background and brief description

[0002] Vehicles contain various components that deteriorate at different rates and require maintenance at different times. Additionally, the deterioration rate of each component can be influenced by several parameters, some of which overlap with other components, while others do not. For example, a system battery in hybrid electric vehicles can deteriorate based on the rate of battery usage, the battery's age, temperature conditions, battery type, and other factors. Similarly, an air filter connected to the engine intake can deteriorate based on the filter's age, air quality, ambient weather conditions, and other factors.

[0003] Several approaches have been developed to predict the health of a vehicle component. One example is presented by Uchida in US 8,676,4825, which predicts the battery health of a hybrid vehicle based on a decrease in the vehicle's fuel efficiency. Another example is presented by Kozlowski et al. in US 20030184307, which predicts the health of a system battery based on the frequency of charging and discharging the battery and its effects on battery parameters such as impedance, electrolyte state, etc. Battery health is then expressed as a number of remaining usage cycles.

[0004] Further relevant prior art is disclosed in US 2016 / 0 112 74 A1 and DE 10 2013 220 691 A1.

[0005] The inventors of the present invention have, however, identified several problems with such approaches. For example, the approaches described above rely on statistical analyses, which can be computationally intensive. Consequently, they may require excessive memory and processor resources to assess the battery's health. Another example is that the approaches described above require frequent measurements from sensors coupled to the relevant components. The reliance on sensors, which are themselves subject to wear and tear, can lead to inaccuracies in the health assessment. Furthermore, the approach does not accurately account for the effect of temperature on the battery's internal resistance and capacity as the battery ages.As another example, an operator may not be able to understand the extent of battery degradation if battery health is indicated by a number of remaining usage cycles. This can be particularly problematic if the battery is part of a hybrid vehicle, where the combustion engine automatically takes over to meet driver demand when the battery is unable to do so. As a result, the operator may be unable to replace or service the battery before it is completely degraded, impacting vehicle operation. Furthermore, the operator may be unable to adjust driving behavior in a timely manner to prevent battery degradation.

[0006] Against this background, the invention relates to methods according to claims 1 and 10. Advantageous embodiments result from the dependent claims.

[0007] In one example, some of the aforementioned problems can be addressed by a procedure for a vehicle that includes: predicting the deterioration state of a vehicle component based on a specific metric derived from a captured vehicle operating parameter, including a past history of that metric; and converting the predicted deterioration state into a remaining time or duration for display to a vehicle operator based on past driving history data and predicted future driving, including the past history of that metric. This allows for a more accurate prediction of the remaining service life of a vehicle component and enables the information to be communicated to the vehicle operator in a more understandable way. In one example, the vehicle component is a system battery.

[0008] As an example, a hybrid vehicle system might include a component whose service life is predicted using statistical methods. A controller can predict a baseline deterioration rate for the component based on a history (e.g., frequency) of the component's maintenance. For example, the controller might use a linear deterioration model to predict a baseline value for the remaining service life of the component. The controller can then update the estimate based on the nature of vehicle operation (e.g., vehicle driving pattern and other driving statistics), the nature of component operation (e.g., how often the component was used in the current driving cycle and in response to which events), and any disturbances or parameters that might change the baseline deterioration rate of that particular component.As an example, if the component being assessed is a battery, the baseline degradation rate can be based on when the battery was last serviced, its state of charge, and temperature conditions. The model can use the measured parameters to estimate the current state of the battery's internal resistance and capacity. The battery health is then calculated as a function of the estimated internal resistance and capacity, with a weighting assigned to the resistance and capacity values ​​varying based on the battery's nature (e.g., whether it is a lead-acid or lithium-ion battery). Furthermore, the battery health can be updated based on how aggressively the vehicle has been driven and on any specific driving maneuvers (e.g.,Turning maneuvers), which quickly discharge the battery. In another example, if the component being assessed is an intake air filter, the baseline deterioration rate can be based on when the filter was last replaced, and the baseline rate can be updated based on detected changes in manifold airflow at different degrees of throttle opening during combustion engine transitions, as well as ambient weather conditions that could cause sudden filter clogging (e.g., the presence of a sudden dust storm or blizzard that could clog the filter). The detected health status can then be converted into an estimate of the component's remaining service life, including a time and / or distance of vehicle travel remaining before the component needs to be replaced or serviced.The conversion can be based on the recorded health status of the filter and further on vehicle driving statistics, including time and / or distance already traveled by the vehicle, as well as the operator's driving patterns and habits. For example, the remaining battery life can be used by the vehicle operator to determine whether a hybrid vehicle can be started using the battery in a given situation.

[0009] In this way, the remaining lifespan of a vehicle component can be accurately predicted without relying on computationally intensive algorithms. By using data collected on board the vehicle, in conjunction with vehicle driving statistics, the health of a component can be calculated more accurately. For example, the internal resistance and capacity of a system battery can be better determined by considering temperature effects as well as the impact of aggressive driving behavior. As another example, the degree of clogging of an air filter can be predicted more accurately based on a recursive estimation of mean and standard deviation of airflow values ​​at larger throttle openings.By assessing an air filter while relying on airflow or manifold pressure data acquired during vehicle transitions, a larger portion of the data collected over a vehicle driving cycle can be used for filter prediction. Additionally, the need to actively maintain the internal combustion engine within a defined RPM / load range to complete a prediction or diagnostic routine is reduced. By converting the acquired health status into an estimate of the remaining time or duration of vehicle operation before component maintenance is required, a vehicle operator can be better informed about the component's condition. Consequently, timely maintenance of the component can be ensured, thereby improving vehicle performance.By predicting the remaining service life of a vehicle component through a recursive estimation of statistical features, the component's remaining service life can be predicted with lower computational intensity without compromising the accuracy of the prediction. This makes it possible to provide a margin that ensures the component's sound operation for the estimated remaining service life.

[0010] The predictive feature can provide an early indication of the component's remaining lifespan, helping customers plan maintenance in advance and prevent component failure. Additionally, the convenience of online estimation can be provided in an easy-to-implement package. Brief description of the drawings Fig. Figure 1 schematically represents an exemplary embodiment of a cylinder of an internal combustion engine coupled in a hybrid vehicle system. Fig. Figure 2 shows exemplary lifetime profiles in percent for a vehicle component operating under different conditions. Fig. Figure 3 shows a high-level flowchart for performing forecasting and diagnosis of a vehicle component using captured data and statistical estimates. Fig. Figure 4 shows a high-level flowchart of an exemplary procedure for applying a statistical estimation to a recorded health status of a vehicle component to predict the remaining service life of the component. Fig. Figure 5 shows an example routine that can be used to predict the remaining lifespan of a vehicle battery. Fig. Figure 6 shows a block diagram of an example algorithm that can be used to predict the remaining lifespan of a vehicle battery. Fig. Figure 7 shows maps illustrating exemplary trends in the health of a vehicle battery's capacity and resistance over time. Fig. Figure 8 shows an exemplary routine that can be used to predict the remaining service life of an intake air filter of an internal combustion engine of a vehicle. Fig. Figure 9 shows a map illustrating an example of the change in manifold airflow through an air filter at different intake throttle angles. Fig. 10 and Fig. Figure 11 shows exemplary experimental data illustrating changes in the mean and standard deviation values ​​of manifold airflow with a change in the throttle angle for filters with different degrees of clogging. Fig. Figure 12 graphically represents a change in the mean and standard deviation values ​​of manifold airflow with a change in the throttle angle for filters with different degrees of clogging. Detailed description

[0011] The following description concerns systems and methods for predicting the remaining service life of a component of a hybrid vehicle system, such as the exemplary vehicle system from Fig. 1. An example controller can be configured to run a control routine, such as the example routines from the Fig. 3 and Fig. 4. To carry out statistical procedures to predict the remaining service life of a vehicle component. The control system can consider the component's deterioration profile as well as the vehicle's driving characteristics to generate percentage service life profiles, such as the exemplary profiles from Fig. 2. The control system can then use the percentage lifetime profile to provide the vehicle operator with the remaining time or distance until the component requires maintenance. Vehicle operations can then be adjusted accordingly. A routine that predicts the remaining lifetime of a vehicle battery (such as the routine from the Fig. 6-7) can rely on detected and predicted changes in the resistance and capacity of the battery, as referenced in Fig. 7 shown. A routine that predicts the remaining service life of an internal combustion engine's intake air filter (such as the routine from Fig. 8) can rely on detected changes in manifold airflow with changing throttle position during transitional conditions, as described in reference to Fig. 9 shown. The controller can then predict the remaining service life of the air filter based on a recursive estimate of the mean and standard deviation of the measured manifold airflow, as shown in the Fig. Figures 10-12 are shown. This way, regular maintenance of the component can be better ensured.

[0012] Fig. Figure 1 represents an example of a combustion chamber or cylinder of an internal combustion engine 10. The internal combustion engine 10 can be coupled to a drive system for driving on a road, such as a vehicle system 5. In one example, the vehicle system 5 could be a hybrid electric vehicle system.

[0013] The engine 10 can be controlled, at least partially, by a control system comprising the control unit 12 and by input from a vehicle operator 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. The cylinder (here also referred to as the "combustion chamber") 14 of the internal combustion engine 10 can have combustion chamber walls 136 in which a piston 138 is arranged. The piston 138 can be coupled to the crankshaft 140, so that a reciprocating motion of the piston is translated into a rotational motion of the crankshaft. The crankshaft 140 can be coupled to at least one drive wheel of the passenger car via a transmission system. Furthermore, a starter (not shown) can be coupled to the crankshaft 140 via a flywheel to enable the engine 10 to be started.

[0014] Cylinder 14 can draw in intake air via a series of intake air ducts 142, 144, and 146. Air drawn in via intake air duct 142 can be filtered by the air filter 135 before moving into air ducts 144 and 146. In addition to cylinder 14, intake air duct 146 can also communicate with other cylinders of the internal combustion engine 10. In some examples, one or more of the intake ducts may include a charging device, such as a turbocharger or a supercharger. For example, [reference to relevant example] shows Fig. 1. The internal combustion engine 10 is configured with a turbocharger comprising a compressor 174 located between the intake ports 142 and 144, and an exhaust turbine 176 located along an exhaust port 148. The compressor 174 can be powered, at least partially, via a shaft 180 through the exhaust turbine 176 when the charging device is configured as a turbocharger. In other examples, such as when the internal combustion engine 10 is equipped with a supercharger, the exhaust turbine 176 can be optionally omitted, with the compressor 174 being driven by mechanical inputs from an electric motor or the internal combustion engine. A throttle 162, which includes a throttle valve 164, can be provided along an intake port of the internal combustion engine to vary the flow rate and / or pressure of the intake air supplied to the internal combustion engine cylinders.For example, the throttle 162 can be positioned downstream of the compressor 174, as shown in . Fig. 1 shown, or alternatively it can be provided upstream of compressor 174.

[0015] The exhaust channel 148 can receive exhaust gases from other cylinders of the internal combustion engine 10 in addition to those from cylinder 14. As shown, the exhaust gas sensor 128 is coupled to the exhaust channel 148 upstream of an emission control device 178. The sensor 128 can be selected from various suitable sensors for providing an indication of an exhaust air-fuel ratio, such as a linear lambda sensor or UEGO sensor (universal exhaust gas oxygen sensor; wide-range or broadband lambda sensor), a dual-state lambda sensor or EGO sensor (as shown), a HEGO sensor (heated EGO sensor), a NOx, HC, or CO sensor. The emission control device 178 can be a three-way catalytic converter (TWC), a NOx trap, various other emission control devices, or combinations thereof.

[0016] Each cylinder of the engine 10 can include one or more intake valves and one or more exhaust valves. For example, cylinder 14, as shown, includes at least one intake control valve 150 and at least one exhaust control valve 156, which are arranged in an upper region of cylinder 14. In some examples, each cylinder of the internal combustion engine 10, including cylinder 14, can include at least two intake control valves and at least two exhaust control valves, which are arranged in an upper region of the cylinder.

[0017] The inlet valve 150 can be controlled by the controller 12 via the actuator 152. Likewise, the exhaust valve 156 can be controlled by the controller 12 via the actuator 154. Under certain conditions, the controller 12 can vary the signals provided to the actuators 152 and 154 to control the opening and closing of the corresponding inlet and exhaust valves. The position of the inlet valve 150 and the exhaust valve 156 can be determined by appropriate valve position sensors (not shown). The valve actuators can be of the electrically actuated type, the cam-actuated type, or a combination thereof. The inlet and exhaust valve actuation can be controlled simultaneously, or any of the following options can be used: variable inlet cam actuation, variable exhaust cam actuation, dual independent variable cam actuation, or fixed cam actuation.Each cam actuation system can include one or more cams and use one or more cam profile switching (CPS), variable cam timing (VCT), variable valve timing (VVT), and / or variable valve lift (VVL) systems, which can be operated by the control unit 12, to vary the valve operation. For example, cylinder 14 can alternatively include an intake valve controlled by an electronic valve actuation system and an exhaust valve controlled by a cam actuation system, including CPS and / or VCT. In other examples, the intake and exhaust valves can be controlled by a common valve actuator or actuation system, or by a variable valve timing actuator or actuation system.

[0018] Cylinder 14 can have a compression ratio that is the volume ratio between piston 138 at bottom dead center and at top dead center. In one example, the compression ratio is in the range of 9:1 to 10:1. However, in some examples where different fuels are used, the compression ratio can be higher. This can occur, for example, when fuels with a higher octane rating or fuels with a higher latent heat of vaporization are used. The compression ratio can also be higher when using direct injection due to its effect on internal combustion engine knock.

[0019] In some examples, each cylinder of the internal combustion engine 10 may include a spark plug 192 to initiate combustion. The ignition system 190 can provide a spark to the combustion chamber 14 via the spark plug 192 in response to a spark advance signal (SA) from the control unit 12 under selected operating modes. However, in some embodiments, the spark plug 192 may be omitted, such as when the engine 10 can initiate combustion by auto-ignition or by fuel injection, as may be the case with some diesel engines.

[0020] In some examples, each cylinder of the internal combustion engine 10 can be configured with one or more fuel injection devices to supply it with fuel. As a non-restrictive example, cylinder 14, as shown, includes two fuel injection devices 166 and 170. The fuel injection devices 166 and 170 can be configured to deliver fuel received from the fuel system 8. The fuel system 8 can include one or more fuel tanks, fuel pumps, and fuel distributors. As shown, the fuel injection device 166 is directly coupled to cylinder 14 to inject fuel directly into it, proportional to the pulse width of the signal FPW-1, which is received by the controller 12 via an electronic driver 168.The fuel injection device 166 provides so-called direct injection (hereinafter referred to as "DI") of fuel into the combustion cylinder 14. While the injection device 166 is shown in the illustration in . Fig. The fuel injector 166 can be positioned on one side of cylinder 14, or alternatively, above the piston, such as near the spark plug 192. Such a position can improve mixing and combustion when the engine is operated with an alcohol-based fuel, as some alcohol-based fuels have lower volatility. Alternatively, the injector can be positioned above and near the intake valve to improve mixing. Fuel can be supplied to the fuel injector 166 from a fuel tank of the fuel system 8 via a high-pressure fuel pump and fuel distributor. Furthermore, the fuel tank can have a pressure converter that provides a signal to the control unit 12.

[0021] The fuel injection device 170 is, as shown, arranged in a configuration that provides so-called port fuel injection (hereinafter referred to as "PFI") into the intake manifold upstream of cylinder 14, in the intake port 146 instead of in cylinder 14. The fuel injection device 170 can inject fuel taken from the fuel system 8 proportionally to the pulse width of the FPW-2 signal received by the controller 12 via the electronic driver 171. It should be noted that a single driver 168 or 171 can be used for both fuel injection systems, or, as shown, several drivers can be used, for example, driver 168 for fuel injection device 166 and driver 171 for fuel injection device 170.

[0022] In an alternative example, each of the fuel injection devices 166 and 170 can be configured as a direct fuel injection device for injecting fuel directly into cylinder 14. In yet another example, each of the fuel injection devices 166 and 170 can be configured as port fuel injection devices for injecting fuel upstream of the intake valve 150. In still further examples, cylinder 14 can contain only a single fuel injection device configured to receive different fuels from the fuel systems in varying relative amounts as a fuel mixture, and further configured to inject this fuel mixture either directly into the cylinder as a direct fuel injection device or upstream of the intake valves as a port fuel injection device.It should therefore be understood that the fuel systems described here are not limited by the specific configurations of fuel injection devices described here as examples.

[0023] Fuel can be supplied to the cylinder by either injection device during a single cylinder cycle. For example, each injection device can provide a portion of the total fuel injection that is burned in cylinder 14. Furthermore, the distribution and / or relative amount of fuel supplied by each injection device can vary with operating conditions, such as combustion engine load, knocking, and exhaust gas temperature, as described below. Fuel injected into the intake manifold can be supplied during an open intake valve event, a closed intake valve event (e.g., essentially before the intake stroke), and during operation with both the open and closed intake valves.Similarly, directly injected fuel can be supplied, for example, during an intake stroke, partially during a preceding exhaust stroke, during the intake stroke, and partially during the compression stroke. Thus, even in a single combustion event, injected fuel can be injected from the intake manifold and direct injection systems at different times. Furthermore, multiple injections of the supplied fuel can be performed per cycle during a single combustion event. These multiple injections can occur during the compression stroke, the intake stroke, or any suitable combination thereof.

[0024] The fuel injection devices 166 and 170 can have different characteristics. These include differences in size; for example, one injection device may have a larger injection orifice than the other. Other differences include, but are not limited to, different spray angles, different operating temperatures, different objectives, different injection timings, different spray characteristics, different positions, etc. Furthermore, different effects can be achieved depending on the distribution ratio of the injected fuel between the injection devices 170 and 166.

[0025] Fuel tanks in fuel system 8 can contain different types of fuel, such as fuels with varying properties and compositions. These differences can include variations in alcohol content, water content, octane rating, heat of vaporization, fuel blends, and / or combinations thereof. For example, fuels with different heats of vaporization could include gasoline as the primary fuel type, with a lower heat of vaporization, and ethanol as the secondary fuel type, with a higher heat of vaporization. Alternatively, the internal combustion engine could use gasoline as the primary fuel type and an alcoholic fuel blend, such as E85 (approximately 85% ethanol and 15% gasoline) or M85 (approximately 85% methanol and 15% gasoline), as the secondary fuel type.Other possible substances include water, methanol, a mixture of alcohol and water, a mixture of water and methanol, a mixture of alcohols, etc.

[0026] In yet another example, both fuels could be alcohol mixtures with varying alcohol compositions. The first fuel could be a gasoline-alcohol mixture with a lower alcohol concentration, such as E10 (which consists of approximately 10% ethanol), while the second fuel could be a gasoline-alcohol mixture with a higher alcohol concentration, such as E85 (which consists of approximately 85% ethanol). Furthermore, the first and second fuels could also differ in other fuel properties, such as temperature, viscosity, octane rating, etc. Additionally, the fuel properties of one or both fuel tanks can change frequently, for example, due to daily fluctuations in refueling.

[0027] Control 12 is in Fig. 1 is represented as a microcomputer comprising a microprocessor unit 106, input / output ports 108, an electronic storage medium for executable programs and calibration values, in this specific example represented as non-volatile read-only memory 110 for storing executable instructions, random access memory 112, keep-alive memory 114 and a data bus.In addition to the signals discussed previously, the control unit 12 can receive various signals from sensors coupled to the engine 10, including the measurement of mass air flow (MAF) from mass air flow sensor 122; air pressure from BP sensor 137; engine coolant temperature (ECT) from temperature sensor 116, which is coupled to cooling sleeve 118; a profile ignition pickup signal (PIP) from Hall effect sensor 120 (or other type), which is coupled to the crankshaft 140; throttle position (TP) from a throttle position sensor; and manifold absolute pressure (MAP) signal from sensor 124. An internal combustion engine speed signal (RPM) can be generated by the control unit 12 from the PIP signal. The manifold pressure signal MAP from a manifold pressure sensor can be used to provide an indication of vacuum or pressure in the intake manifold.The controller 12 receives signals from the various sensors. Fig. 1 and suspends the various actuators Fig. 1. To adjust the combustion engine operation based on received signals and instructions stored in the controller's memory. For example, based on a pulse width signal commanded by the controller to a driver coupled to the direct injection device, a fuel pulse can be delivered by the direct injection device to a corresponding cylinder. Exemplary routines that can be executed by the controller are related to the Fig. Shown in 3-5 and 8.

[0028] As described above, shows Fig. 1 merely one cylinder of a multi-cylinder internal combustion engine. Thus, each cylinder can likewise have its own set of intake / exhaust valves, fuel injection device(s), spark plug, etc. It is understood that the internal combustion engine 10 can include any suitable number of cylinders, including 2, 3, 4, 5, 6, 8, 10, 12 or more cylinders. Furthermore, each of these cylinders can contain some or all of the various components that are in Fig. 1 are described and illustrated with reference to cylinder 14.

[0029] In some examples, the vehicle 5 can be a hybrid vehicle with multiple torque sources available to one or more vehicle wheels 55. In other examples, the vehicle 5 is a conventional vehicle with only an internal combustion engine or an electric vehicle with only one electric machine. In the example shown, the vehicle 5 includes an internal combustion engine 10 and an electric machine 52. The electric machine 52 can be an electric motor or an electric motor / generator. The crankshaft 140 of the internal combustion engine 10 and the electric machine 52 are connected to the vehicle wheels 55 via a transmission 54 when one or more clutches 56 are engaged. In the example shown, a first clutch 56 is provided between the crankshaft 140 and the electric machine 52, and a second clutch 56 is provided between the electric machine 52 and the transmission 54.The control unit 12 can send a signal to an actuator of each clutch 56 to engage or disengage the clutch, thereby connecting or disconnecting the crankshaft 140 from the electric machine 52 and its associated components, and / or connecting or disconnecting the electric machine 52 from the transmission 54 and its associated components. The transmission 54 can be a manual transmission, a planetary gear system, or another type of transmission. The powertrain can be configured in various ways, including as a parallel, in-line, or in-line-parallel hybrid vehicle.

[0030] The electric machine 52 receives electrical power from a traction battery 58 to provide torque to the vehicle wheels 55. The electric machine 52 can also be operated as a generator to provide electrical power for charging the battery 58, for example, during braking. It may be necessary to regularly maintain and diagnose the various vehicle components. Additionally, the remaining service life of each component can vary based on its maintenance or deterioration history. For example, the controller can periodically diagnose and recursively estimate the remaining service life of the vehicle system battery, the intake air filter (Inventor: please insert examples of other components to be diagnosed). As with reference to the Fig. 3-5 and 8, a vehicle control unit can be configured to use an algorithm to make a statistical prediction regarding the remaining service life of the component based on a previous history of the component's deterioration behavior, on collected data for parameters relating to the component, and on mapped vehicle driving statistics (such as real-time vehicle driving statistics or those compiled over a current vehicle driving cycle). The control unit can then provide the vehicle operator with useful information regarding the component, such as an estimate of the distance or time until maintenance.

[0031] The components from Fig. 1. Activate a vehicle system comprising an internal combustion engine; an electric motor powered by a battery; sensors for measuring battery voltage and current; and a controller. The controller can be configured with computer-readable instructions stored in non-volatile memory for the following: predicting a battery deterioration state based on the specified battery capacity and resistance, derived from the detected battery current or voltage, including a past history of the specified battery resistance and capacity. The controller can further include instructions to convert the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator, based on past driving history data and predicted future driving, including the past history of the specified metric.

[0032] The components from Fig. 1 further activate a vehicle system comprising an internal combustion engine, including an intake manifold; an air filter coupled to the intake manifold; an intake throttle; a manifold airflow sensor coupled downstream of the intake throttle; and a control system.The controller can be configured with computer-readable instructions stored on non-volatile memory for the following: storing the measured airflow readings when the intake throttle is commanded to exceed a threshold throttle angle; estimating a metric that specifies a distribution of manifold airflow based on the stored measured airflow readings; predicting an air filter deterioration state based on the estimated metric relative to a threshold; and converting the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator based on past driving history data and predicted future driving, including a past history of the estimated metric. As an example, the prediction may include forecasting a higher state of deterioration if the estimated metric falls below the threshold.The metric can be a primary metric, and the control can further include instructions for estimating a secondary metric indicating average manifold airflow through the air filter. In this case, the prediction can include predictions of the higher state of deterioration when the secondary metric falls below the threshold. The threshold can be determined as a function of a recent estimate of the metric, retrieved from the earlier history of the estimated metric, and a distance traveled by the vehicle since the most recent estimate. Alternatively, the threshold can be determined as a function of an initial estimate of the metric at the time of the air filter's installation, retrieved from the earlier history of the estimated metric, and a distance traveled by the vehicle since the air filter's installation. With reference to... Fig. Figures 200, 220, and 230 represent exemplary deterioration models for a vehicle component. In each case, the x-axis represents time and the y-axis represents the percentage of remaining service life, where T100 represents the time when the component reaches the end of its service life L100.

[0033] It is assumed that the deterioration behavior of a component is generally constant throughout its lifetime. Assuming constant deterioration behavior, as shown with reference to Chart 200, a linear deterioration model can be determined, for example, where the remaining lifetime is calculated according to the following equation: L = alpha * T, where alpha is the deterioration rate. At an intermediate time (tinterm) before component failure, the component can be predicted with a percentage lifetime as a linterm. Here, tinterm is the time elapsed between the start of the component's lifetime (L0, when the component is first put into operation, such as after a replacement, maintenance, reset, etc.) and the time when the prediction is made in real time.Therefore, based on the model, the linterm can be determined as follows: Linterm = alpha*Tinterm. The remaining time (TRL) until the end of the component's lifetime can be determined as follows: TRL=(T100−Tinterm)=(L100−Linterm) / alpha

[0034] Alpha is estimated in real time based on the component's degradation (past observation, between T0 and the tinterm). L100 is provided as an input to define the component's end of life. The tinterm is then estimated in real time using the prediction algorithm.

[0035] However, due to the nature of the component's operation and its interaction with disturbances, conditions may arise that cause a sudden change in the component's remaining lifetime. For example, the intake air filter may slowly become clogged with a linear deterioration rate due to dust accumulation on the filter during driving and the intake of combustion engine air. However, while driving the vehicle through snow or in the presence of dusty conditions (such as driving through a sandstorm), the air filter may become more heavily clogged in a shorter time, with the additional clogging being clear or unclear depending on the type of accumulation. In such a case, modeling the remaining lifetime using a constant parameter can still assume a constant deterioration rate based on past statistics and adjust the rate in real time based on a real-time assessment of the conditions.

[0036] For example, the model, referring to map 220, can initially assume a constant deterioration rate between T0 and the tinterm. The component at the tinterm may undergo a drastic change, which alters the estimate of the percentage remaining lifetime from linterm to L'interm, where L'interm = linterm + deltaL. Assuming a constant deterioration rate between T0 and T'interm (as between T0 and the tinterm, since linterm = alpha*tinterm), the remaining lifetime is calculated as follows: TRL=(T100−Tinterm)=(L100−Linterm−deltaL) / alpha=(L100−L'interm) / alpha

[0037] It is understood that the drastic change can be positive or negative, and the equation can apply in both scenarios.

[0038] Referring to map 230, the model can initially assume a constant deterioration rate between T0 and Tinterm1. At Tinterm1, a first, positive drastic change can occur, necessitating a change in the estimate of the remaining percentage lifetime from Linterm to L'interm. A constant deterioration rate can then be assumed between Tinterm1 and Tinterm2. At Tinterm2, a second, negative drastic change can occur, requiring a reversion of the estimate of the remaining percentage lifetime from L'interm to Linterm. Afterward, a constant deterioration rate can again be assumed up to L100.As an example of such a condition, if an air filter is partially clogged due to snow, there can be a drastic change in the filter's condition, thus extending its remaining service life closer to the end of its lifespan. However, if the snow melts or the filter is removed, cleaned, and reinstalled, there is another drastic change in the filter's condition, extending its remaining service life.

[0039] With reference to Fig. Figure 3 shows an exemplary method 300 for estimating the remaining service life of a vehicle component. Instructions for executing method 300 and the remaining methods included herein can be executed by a controller based on instructions stored in a memory of the controller and in conjunction with signals received from sensors of the internal combustion engine system, such as those described above with reference to Fig. The control system can utilize actuators of the vehicle system and the internal combustion engine system to diagnose the component's health status according to the procedures described below. This procedure enables the prediction of a deteriorating vehicle component based on the change in the value of a metric associated with the vehicle component from an initial value of the metric at the time of its installation in the vehicle system over the course of vehicle operation. The prediction is further based on the distance traveled by the vehicle over this period, with the metric being derived from a recorded vehicle operating parameter.The method further enables the conversion of the predicted state of deterioration into an estimate of the time or duration remaining before the vehicle component needs to be serviced (or replaced) for display to a vehicle operator, the conversion being based on any of the previous driving history data and predicted future driving.

[0040] In case 302, the procedure involves estimating and / or measuring vehicle operating conditions. These may include, for example, vehicle speed, internal combustion engine speed, pedal position, driver-demanded torque, environmental conditions (such as ambient temperature, pressure, and humidity), boost pressure, battery charge status, manifold airflow, exhaust air-fuel ratio, transmission gear selection, driving mode (e.g., electric or internal combustion engine mode; sport, performance, or economy mode), etc.

[0041] In section 303, the procedure involves selecting a component for evaluation. A component can be evaluated periodically. In this case, the selection of a component can be based on the time or distance the vehicle has traveled since the component was last evaluated. In another example, the selection can be based on an active request received from the operator. This can be in addition to or independent of the periodic evaluation. For example, an operator might request a system battery forecast before beginning a planned route.

[0042] For procedure 304, this involves retrieving the deterioration or maintenance history of the component to be assessed. This includes retrieving a time or duration elapsed since the component was first installed or operated in the vehicle. Alternatively, it may involve retrieving a time or duration elapsed since the component was last serviced, repaired, or reset.

[0043] Additionally, the maintenance history may include details of a component deterioration rate prior to the most recent maintenance event, a baseline component deterioration rate, an average component deterioration rate over the vehicle's lifetime, and any diagnostic code associated with the component that has been activated over the vehicle's lifetime.

[0044] In section 306, the procedure involves using data collected on board the vehicle to determine the component's health status. As referenced in the examples from the Fig. 5 and Fig. As described in section 8, this involves acquiring one or more parameters associated with the component and comparing the acquired data for a current iteration of the routine with data acquired in a previous iteration to update the component's deterioration rate (from a baseline rate) in real time. For example, the controller can predict the deterioration state of a vehicle component based on a specific metric derived from an acquired vehicle operating parameter, including a previous history of that metric. In one example, the vehicle component is a system battery, the specific metric is one or more of a battery resistance and battery capacity, and the acquired vehicle operating parameter includes one or more of a battery current and battery voltage.In another example, the vehicle component is an intake air filter of the internal combustion engine, and the defined metric is one or more of a mean and a standard deviation value of airflow through the filter. The measured vehicle operating parameter includes manifold airflow. In each case, the measured vehicle operating parameter is selected based on the vehicle component to be predicted.

[0045] From 306, the procedure proceeds to 308, where the component is diagnosed based on its health status (as determined from the collected data) relative to a threshold. For example, if it is determined that the estimated health status falls below a component-specific threshold, the component may be considered deteriorated, and an indication may be provided to the operator that the component needs to be repaired, serviced, or replaced. The procedure then proceeds from 306 to 310, where the vehicle driving statistics are retrieved. The vehicle driving statistics may include, for example, a distance traveled over the vehicle's lifetime (e.g., based on an odometer reading), the number and frequency of maintenance events that have occurred over the vehicle's lifetime (e.g.,Vehicle driving statistics include data such as the number of oil changes, their frequency, and the odometer reading at which they were performed; the vehicle's average fuel efficiency; average vehicle speed; average gear usage; average daily mileage; average tire pressure; and more. These statistics can also include operator-specific driving patterns and habits. For example, this might include an operator's preference for fuel efficiency over performance, the frequency and degree of pedal use and actuation (e.g., whether the operator has a heavy foot), the operator's driving style, and the average speed at which the operator drives.Vehicle driving statistics can also include details regarding the weather conditions in which the vehicle is typically driven, such as whether the vehicle is typically operated in rain or snow, dry or wet conditions, etc. Vehicle driving characteristics can reflect operator driving tendencies and average conditions experienced by the vehicle component, which can influence the component's initial rate of deterioration.

[0046] In 312, the estimated component health (based on the collected data) and the retrieved vehicle driving statistics can be used in combination to perform a recursive estimation of the time or distance remaining before the component deteriorates. For example, the estimated component health can be updated to reflect the vehicle driving statistics, and then the updated health can be converted into an estimate of the time / distance remaining before the component deteriorates. In one example, the controller can use an algorithm, such as the example algorithm from Fig. 6 or the exemplary routine from Fig. 8, which are based on the algorithm from Fig. 4 is combined to convert the updated health status into an estimate of the time / distance remaining before component deterioration occurs. For example, the controller can convert the predicted deterioration status into an estimate of the remaining time or duration for display to a vehicle operator, based on past driving history data and predicted future driving, including the past history of the specified metric. Additionally, the predicted deterioration status can be converted into a remaining number of fuel refueling events for display to the vehicle operator, based on past driving history data and predicted future driving.

[0047] At 314, the control unit can display the estimated time / distance remaining before component deterioration to the vehicle operator, such as on a display screen in the vehicle's center console.

[0048] With reference to Fig. Figure 4 shows an exemplary procedure 400 for estimating information regarding time or distance until maintenance for a vehicle component. The algorithm is based on the statistical characterization of a driving frequency or duration as a function of absolute time. The prediction of the remaining distance (or time) is based on a previous deterioration rate of the component together with the vehicle's driving statistics. The distance estimate can be determined as the product of the remaining time before the component deteriorates (in days) and the distance driven by the vehicle per day. The accuracy of the distance estimate can be further improved by taking into account driving variations (e.g.,Driving deviations for specific days of the week or deviations between weeks) and further improved by taking into account a safety margin that better ensures fault-free operation of the component during the estimated remaining service life.

[0049] For example, at any given time, the algorithm can be based on a deterioration rate model to increase the standard deviation as follows: %Deterioration(t) = μDeterioration(t) + n*σDeterioration(t) where n is the safety factor, which can be calibrated, and where mu Verschlechterung (t) and sigma Verschlechterung (t) are the real-time estimates of the deterioration model.

[0050] In procedure 402, this involves determining that the internal combustion engine is running. For example, if the vehicle is a hybrid, it can be determined whether the vehicle is operating in an internal combustion engine mode or an auxiliary mode, with at least some of the torque requirement being met by the internal combustion engine torque. If the internal combustion engine is not running, the procedure can end. Since there is no airflow through the air filter in electric mode, no deterioration is expected. The same applies to other gasoline-internal combustion engine components (such as an oil filter, coolant pump, etc.). Other components, such as a battery and alternator, are still functional during electric operation. Therefore, they can still be monitored / predicted while the internal combustion engine is off.

[0051] In 406, the procedure involves incrementing a timer to provide a real-time estimate of the elapsed time (t). Subsequently, in 408, the procedure involves checking that the last run of the algorithm was not performed on the current day on which the routine is being executed ("today"). In 410, the procedure involves incrementing the distance traveled at time t as follows: dist(t)=dist(t−1)+Vspd*Δt, where dist (t) is the distance covered at time t, dist (t-1) is the distance covered at the last repetition of the routine (t-1), Vspd is the vehicle speed and Δt is the time that has elapsed since the last repetition of the routine.

[0052] In version 412, the procedure optionally includes recursively updating one or more statistical parameters of the data collected for the assessed component. These parameters include the mean of the collected data (µtoday(t)) and the standard deviation of the collected data (σtoday(t)). The updated values ​​are then stored as a function of the previous values, which are stored at the time of the last iteration of the routine (e.g., last week) and the distance covered (dist(t)). Depending on how frequently the algorithm is expected to report / update the remaining lifespan of the component being monitored / predicted, the routine is updated either daily or within the same day if the target component is prone to rapid deterioration.

[0053] Referring again to 408, if the current iteration of the algorithm is performed on the fixed day (“today”) selected for running the routine, then 420 may specify that the day of the last iteration of the routine is the current day (of today). 422 optionally includes the recursive updating of one or more statistical parameters of the data collected for the evaluated component, where the one or more parameters include a mean of the collected data (µtoday-1(t)) and a standard deviation of the collected data (σtoday-1(t)). The updated values ​​are then stored as a function of the previous values ​​stored at the time of the last iteration of the routine (e.g., last week) and the distance covered (dist(t)), i.e., the estimates of µtoday-1 and σtoday-1 for last week and dist(t).Here, mu and sigma estimates are updated based on the old values ​​(from the last run) and the distance traveled on the last day.

[0054] In the 424 case, the procedure involves setting the covered distance to zero to reinitialize the distance counter for the new day. That is, the controller can set dist(t) = 0. For example, suppose the driving statistics are for a specific day, such as Sunday. To update the statistics for Sunday of the current week (that is, the signal and mu values ​​for Sunday), the algorithm can proceed in one of two ways. In the first option, the controller can estimate or update the values ​​in real time (i.e., continuously) and store the result each time new data is collected. Here, sigma(d, w-1), mu(d, w-1), and dist(d, t) can be the inputs. The generated outputs are sigma(d, w) and mu(d, w).In the second option, the controller can continue collecting driving data on Sunday and wait until Monday to update the driving statistics for Sunday once (i.e., in discrete events) and save the new result. Here, sigma(d-1, w-1), mu(d-1, w-1), and dist(d, t) can be the inputs. The generated outputs are sigma(d-1, w) and mu(d-1, w).

[0055] In this way, the algorithm relies on Fig. 4. The statistical characterization of a component's health status as a function of time is used. By predicting the remaining service life based on the previous deterioration rate and performing a linear or non-linear approximation to estimate the component's remaining time, a more reliable predictive approach is provided. Mapping the time to maintenance into distance information using driving statistics also allows the information to be communicated to the vehicle operator in a format that enables timely maintenance of the component and, optionally, adjustments to driving patterns / habits.

[0056] With reference to Fig. Figure 5 shows an exemplary routine 500 for accurately estimating the remaining lifespan of a vehicle system battery. Fig. 6 displays the algorithm Fig. Figure 5 is presented as a block diagram. The method diagnoses and predicts the health of an automotive battery. It employs a high-level, machine learning approach applicable to all types of automotive batteries, including lead-acid and lithium-ion batteries. The equivalent circuit parameters are identified regularly at fixed intervals and are assumed to be explicitly dependent on the battery temperature (θ) and state of charge (SOC).

[0057] The state of health (SOH) of a battery can be expressed as a percentage of its remaining lifespan, ranging from 100% for new batteries to 0% for depleted batteries. As the battery ages and its SOH deteriorates, its internal resistance increases and its internal capacity decreases. The prediction algorithm from the Fig. 5-6 uses a weighted expression of the SOH, which takes into account these increases in R and decreases in C according to the following phenomenon: SOH = SOH R * SOH C

[0058] The internal R and C values ​​are related to the estimated R0, R1, and C1 values ​​that are monitored, and this is how the SOH is estimated. Map 700 from Fig. Figure 7 represents the change in the internal C of a battery over time, while the 750 card is made up of Fig. 7 represents the change in the internal resistance of the battery over time (inventor: please elaborate).

[0059] Regarding procedure 500 from Fig. In 502, the procedure involves estimating and / or measuring temperature conditions. These include, for example, ambient temperature, battery temperature, etc. In 504, the procedure involves retrieving the battery's maintenance history to determine when the battery was last serviced. The last battery service may include battery replacement, repair, or resetting. For example, the duration or distance traveled since the last service can be retrieved. Additionally, the battery's deterioration rate at the time of the last service, as well as the nature / reason for the deterioration, can be retrieved. For example, it can be determined whether the battery deteriorated due to a temperature problem (e.g., overheating), aging, a higher-than-expected wear rate, a vehicle incident / accident, etc.has worsened.

[0060] In the 506, the procedure involves measuring battery parameters in real time. For example, a real-time estimate of the battery current (I) and battery voltage (V) can be performed. The value of the real-time estimate can be compared with reference values ​​of the parameters. As elaborated here, the controller can determine a deterioration state of a vehicle component based on a specific metric derived from a captured vehicle operating parameter, including a previous history of that metric, and convert the predicted deterioration state into a remaining time or duration for display to a vehicle operator based on past driving history data and predicted future driving, including the previous history of that metric.In the example shown, the component is a system battery, the specific metric being one or more of a battery resistance and a battery capacity, and the acquired vehicle operating parameter including one or more of a battery current and a battery voltage. Measuring the battery parameters in real time involves measuring them during vehicle operation, as vehicle and combustion engine operating conditions change. For example, the vehicle operating parameters can be acquired during transient and steady-state vehicle operating conditions. The controller can weight the vehicle operating parameter acquired during transient vehicle operating conditions differently (e.g., more highly than) the vehicle operating parameter acquired during steady-state vehicle operating conditions.

[0061] In the 508, the procedure involves updating a thermal model of the battery based on the measured temperature conditions. For example, the equivalent circuit parameters can be normalized with respect to a reference temperature. As a result, the effect of temperature change on the variation of the equivalent circuit parameters is reduced (in one example, eliminated) as the battery ages. Thus, reference curves in Fig. 7. One-dimensional and independent of temperature. At 510, the method involves using the updated thermal model to predict changes in the battery's resistance and capacity. For example, from Fig. As can be seen in section 7, it can be expected that the normalized internal resistance and the normalized internal capacitance will change when the health of the battery changes.

[0062] In 512, the method involves predicting the battery state of charge (SOC), partly based on recursive estimation of the equivalent circuit model. For example, a higher state of battery deterioration can be predicted if the battery resistance increases or if the battery capacity decreases. In 514, the method involves recursively estimating the battery health based on the predicted state of charge, resistance, and capacity. For example, as the battery ages, the internal resistance is expected to increase, while its capacity decreases. As an example, the estimated battery health can include an estimate of the percentage of life used. For example, if the health is 60%, this indicates that 60% of the battery's lifespan has elapsed, and only 40% of the battery's lifespan remains available for use.

[0063] As an example, the controller can estimate a value of a specific metric (here, battery resistance and capacity) as a function of a recent estimate of the specific metric, retrieved from its previous history, and a distance the vehicle has traveled since the most recent estimate. Furthermore, the controller can estimate a value of the specific metric as a function of an initial estimate of the specific metric, retrieved from its previous history at a point in time when the component was installed in the vehicle.The past driving pattern and the predicted future driving pattern may include one or more of a measure of driving aggressiveness, a rate of pedal usage, frequent driving time patterns, habitual probability patterns, route-based statistical profiles, and environmental attribute profiles, with the environmental attribute profiles including past and predicted weather events along the vehicle's driving route.

[0064] In the case of 516, the procedure involves updating the battery health status based on vehicle driving statistics, including operator driving patterns and habits. Operator driving patterns and habits can include driving patterns and habits retrieved from past driving history data as well as predicted future driving data. The algorithm using the battery health status estimate can rely on a fixed degradation rate. However, as with respect to Fig. As explained in section 2, events may occur that cause the deterioration rate to increase or decrease from the fixed rate. Some of these events may occur outside the vehicle's internal combustion engine, such as weather events, including precipitation (which can affect the component due to an increase in ambient humidity). Other events may involve operator driving patterns, such as the aggressiveness with which the operator tends to drive when driving at a constant speed as opposed to frequently operating the accelerator and brake pedals, the average speed at which the vehicle is driven, the average mode in which the vehicle is driven (e.g., internal combustion engine or electric mode in a hybrid vehicle), etc. Furthermore, a future (e.g.,Predicted driving patterns are taken into account, such as terrain, ambient altitude and temperature, predicted pedal events, and gradients expected along the selected navigation route. For example, if the operator frequently uses the accelerator and brake pedals (or if this is expected given the selected route), the battery may discharge more quickly. Similarly, if the operator drives aggressively, the battery may heat up more quickly and wear out due to higher average battery temperatures. Referring back to the previous example, if the initial estimated battery health based on driving statistics is 60%, the health may be updated to 68%, indicating that the battery is degrading more rapidly, or is expected to degrade more rapidly, due to the vehicle's driving statistics.

[0065] In the case of 518, the procedure involves converting the estimated state of health into an estimate of the remaining battery life, with the remaining life estimate being provided as a time or distance remaining before the battery requires servicing. Specifically, the controller can convert the predicted state of deterioration into an estimate of the remaining time or duration for display to a vehicle operator, the conversion being based on past driving history data and predicted future driving, including the past history of the specified metric (here, the past history of the battery's resistance and capacity profile). The controller can use the estimated state of health and the estimated current vehicle distance (such as that indicated by the vehicle's odometer) to calculate the remaining life.Furthermore, the conversion can be based on an absolute distance traveled by the vehicle, retrieved from previous driving history data.

[0066] At 520, the control unit can display the estimated time / distance remaining until component degradation to the vehicle operator, for example, on a display in the vehicle's center console. It can display, for instance, that "the battery needs to be replaced in 120 miles." This can provide the operator with a more understandable estimate of when the battery maintenance will be required. Furthermore, the displayed estimate can prompt the operator to adjust their driving style, for example, by encouraging less aggressive driving. Additionally, the control unit can convert the predicted degradation state into a remaining number of refueling events for display to the operator, based on past driving history and predicted future driving behavior.

[0067] In this way, the procedure takes into account Fig. 5 explicitly addresses the effects of battery temperature θ and SOC on identifying the state of health and creates 3D (discrete domain) maps of the RC parameters as a function of θ and the SOC. Furthermore, the proposed discrete domain map allows for simpler physical constraints and can relate to different timescales of the system. The strategy is essentially an element-based approach where the RC parameters are identified "periodically" based on the battery lifetime to provide a SOH and time estimate. Consider an nth-order Randel model shown below (a first-order model would have only one RC element):

[0068] The equivalent circuit parameters are estimated in a "fast" manner as a function of temperature and state of charge (SOC). Extensions to a higher-order Randel model are possible. The obtained R0(T, SOC), R1(T, SOC), and C1(T, SOC) are monitored with battery aging to provide an estimate of the remaining service life and its change over time.

[0069] Kirchhoff's rules, applied to the equivalent circuit model, yield the following: {dvcdt=−vcR1(θ)+lC1(θ)v=R0(θ)l+vc where εk is the error measurement, λ is a negligibility factor and γ is a gain matrix.

[0070] Transforming the above equations into the Laplace domain yields: {sVc(s)=−Vc(s)R1(θ)+I(s)C1(θ)V(s)=R0(θ)I(s)+Vc(s)

[0071] The corresponding transfer function is defined as: H(s,θ)=V(s)I(s)=(R0(θ)+R1(θ))+R0(θ)R1(θ)C1(θ)s1+R1(θ)C1(θ)s

[0072] Using the Tustin rule (s←2T2−12+1), The corresponding discrete time transfer function can then be expressed as follows: H(z,θ)=b0(θ)+b1(θ)z−11+a1(θ)z−1where{a1(θ)=1−2R1(θ)C1(θ)1+2R1(θ)C1(θ)b0(θ)=T(R0(θ)+R 1(θ))+2R0(θ)R1(θ)C1(θ)1+2R1(θ)C1(θ)b1(θ)=T(R0(θ)+R1(θ))−2R0(θ)R1(θ)C1(θ)1+2R1(θ)C1(θ) H(z,θ)=V(z)I(z)=b0(θ)+b1(θ)z−11+a1(θ)z−1→V(k)=b0(θ)I(k)+b1(θ)I(k−1)−a1(θ)V(k−1) V(k)=Φxwobeix=[b0(θ)b1(θ)a1(θ)]TΦ=[I(k)I(k−1)V(k−1)]

[0073] A recursive model can then be used to minimize the error between the predicted and the measured voltage as follows: xk=xk−1+γλkεkΦkT where ε k is the error measure, λ is a forgetting factor and γ is a gain matrix

[0074] In this way, a two-way map allows the transformation between the equivalent circuit parameters of the discrete and physical domains. Physical constraints must then be imposed to ensure that R>0 and C>0. The updated formula is then an "approximation" in that it imposes constraints on the size of the update at each time sample.

[0075] The symmetry of the formula is noteworthy.

[0076] The online identification is then repeated for different entry conditions of θ and SOC. A 3D map for the equivalent circuit parameters is then created as follows: R0=f1(θ,SOC) R1=f2(θ,SOC) C1=f3(θ,SOC)

[0077] The initial parameters (R0(0), R1(0) and C1(0)) can be obtained through offline estimation. Crank characteristics and observed system time constraints can also be used to initiate the algorithm. {Vk1=b0(θ)Ik1+b1(θ)Ik1−1−a1(θ)Vk1−1Vk2=b0(θ)Ik2+b1(θ)Ik2−1−a1(θ)Vk2−1Vk3=b0(θ)Ik3+b1(θ)Ik3−1−a1(θ)Vk3−1

[0078] A closed-form formula then yields a1, b0, and b1, where k1, k2, and k3 are discrete time instances to be selected during the initial discharge phase. In this way, a controller can detect the deterioration state of a vehicle battery based on changes in battery resistance and capacity from initial values ​​estimated at a specific point in time within the vehicle system, over a period of vehicle travel, and further based on the distance traveled by the vehicle over that period, where the battery resistance and capacity are derived from a detected battery current or voltage.The control system can then convert the predicted degradation state into an estimate of the time or duration remaining before battery maintenance is required, for display to a vehicle operator. This conversion is based on any previous driving history data and predicted future driving. Here, the battery voltage or current measured during transient vehicle operating conditions can be weighted differently than the battery voltage or current measured during steady-state vehicle operating conditions. For example, the battery voltage or current measured during transient vehicle operating conditions can be weighted more heavily than the battery voltage or current measured during steady-state vehicle operating conditions.

[0079] Fig. Section 6 presents the algorithm of the procedure. Fig. Figure 5 is presented as a block diagram. Block diagram 600 illustrates the connection between the different components of the battery state of health (SOH) estimation. The inputs of the first block, namely battery identification, are the current and voltage. Through online estimation, this component updates the parameters of the equivalent circuit model, namely R0, R1, and C1, under appropriate input conditions. The obtained values ​​are then normalized with respect to reference values ​​corresponding to a reference temperature using the thermal model block. The explicit dependence on the state of charge (SOC) (or equivalent depth of discharge) is ensured by the real-time estimation of the SOC and the open-circuit voltage Em using the "Charge and Capacity" and "Em Calculation" blocks.The outputs from these various components are then fed into the parameter development block, which allows the recording of changes in the parameters of the equivalent circuit model as the battery ages. Together with the trend, the input from the "Driving Statistics" block allows for the prediction and generation of an estimate regarding the remaining battery lifespan. The algorithm includes a battery circuit equation model that allows the prediction of the battery's voltage output for any given input current using the estimated values ​​of Em, R0, R1, and C1.

[0080] In this way, a controller can predict the deterioration state of a vehicle component based on a change in a metric associated with the vehicle component over a period of time and a distance traveled by the vehicle during that period, where the metric is derived from a captured vehicle operating parameter. The controller can then convert the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator, based on past driving history data and predicted future driving. The controller can then display the estimate of remaining time or duration to the vehicle operator as a number of refueling events remaining until the vehicle component requires maintenance.Past driving behavior and predicted future driving behavior can include one or more measures of driving aggressiveness, pedal usage rate, frequent trip time patterns, habitual probability patterns, route-based statistical profiles, and environmental attribute profiles, where environmental attribute profiles include past and predicted weather events along a vehicle's route. Conversion based on distance traveled involves conversion by a factor based on a current vehicle odometer reading. If the vehicle component is a system battery, as in the illustrated case, the captured vehicle operating parameter includes battery current and battery voltage. If the vehicle component is an internal combustion engine intake air filter, as in [reference missing], the [reference missing] Fig. As described in section 8, the recorded vehicle operating parameter includes a manifold airflow and a manifold air pressure. The change in the metric over time can involve a change from an initial value of the metric, estimated at the installation time of the vehicle component. The prediction can involve increasing the deterioration state of the vehicle component toward a fully deteriorated state as the difference between a current value of the metric relative to the initial value of the metric increases above a threshold, where the threshold is based on the distance traveled by the vehicle over time.

[0081] In relation to Fig. Figure 8 presents an exemplary method 800 for estimating the remaining service life of an internal combustion engine intake air filter based on manifold airflow and / or manifold air pressure measurements. This method enables the vehicle operator to be informed of the time or distance of vehicle travel remaining until the air filter requires service / replacement.

[0082] In 802, the procedure involves estimating and / or measuring engine conditions. This may include, for example, combustion engine temperature conditions such as the combustion engine temperature, air temperature, ambient temperature, etc. In 804, the procedure involves retrieving the air filter maintenance history to determine when the air filter was last serviced. The last air filter maintenance may include the air filter being replaced, repaired, or reset. For example, the duration or distance traveled since the last air filter maintenance may be retrieved. Additionally, the rate of deterioration of the air filter at the last maintenance, as well as the nature / reason for the deterioration at the last maintenance, may be retrieved.For example, it can be determined whether the air filter has deteriorated due to aging, a wear rate that is higher than expected, poor quality, a vehicle incident / accident, etc.

[0083] In procedure 806, this involves measuring a change in manifold airflow (MAF) and / or manifold pressure (MAP) when the throttle position is changed during vehicle operation, including both transient and steady-state operating conditions. Specifically, as operator torque demands vary, the control unit monitors the variation in the commanded throttle angle relative to the measured MAF or MAP to assess pressure loss resulting from increasing air filter clogging. Here, the vehicle component being assessed is an internal combustion engine intake air filter; a metric associated with and determined by the filter is one or more of a mean and standard deviation value of the airflow through the filter; and the captured vehicle operating parameter on which the metric is based includes manifold airflow.The control system can give greater weight to manifold airflow detected at higher intake throttle angles than to manifold airflow detected at lower throttle angles. Alternatively, the control system can give greater weight to the vehicle operating parameter detected during transient vehicle operating conditions than to the vehicle operating parameter detected during steady-state vehicle operating conditions. In some examples, the vehicle operating parameter to be detected may include manifold air pressure relative to atmospheric pressure, in addition to or instead of MAF. In 808, the procedure involves confirming whether sufficient data has been collected at a throttle angle above a threshold. That is, health prediction can only be performed after a threshold amount of airflow data has been collected, with an intake throttle at a throttle angle above an upper threshold.For example, it can be confirmed that sufficient data was collected at throttle angles of 55 degrees or more. By giving greater weight to the data collected at larger throttle angles, the effect of the throttle state on the MAF or MAP measurement can be reduced. At larger throttle angles, the effect of the throttle valve as a flow restriction is largely reduced. The expected effect of increased air filter clogging is a decrease in the measured mean MAF or MAP at large throttle angles and a decrease in the corresponding standard deviation. As air filter clogging increases, rapid transitions of the air filter's MAF or MAP become more difficult. Consequently, filter clogging can be better determined in the presence of transients by relying on data collected at larger throttle angles.Furthermore, the need for stationary vehicle operation to evaluate the air filter is reduced.

[0084] If insufficient data has been collected, the procedure continues collecting data at 810. If sufficient data has been collected, such as when more than a calibratable amount of data has been gathered at the larger throttle angle, the procedure proceeds to 812, where the controller recursively estimates a mean (mu or µ) and a standard deviation (sigma or σ) value of the collected MAF data. The mean represents an average value of the airflow through the air filter, while the standard deviation value represents a distribution of the collected airflow measurements.

[0085] At 814, it can be determined whether the estimated mean is below a threshold and / or whether the estimated standard deviation of MAF / MAP is below a threshold at larger throttle angles. Alternatively, the estimated mean and standard deviation values ​​can be compared to expected values. The expected values ​​of the airflow measurements can be based on any distance traveled by the vehicle, previous vehicle driving data, predicted future vehicle driving, and an initial airflow measurement at the time the filter was installed in the vehicle. For example, the mu threshold is a value indicating the maximum airflow or average manifold absolute pressure. Similarly, the sigma threshold is a value indicating the ease with which transients pass through the air filter.If either the mean or standard deviation value is below the corresponding threshold, it can be indicated at 816 that the air filter is deteriorating and requires service or replacement. In this way, the air filter can be diagnosed based on the measured mu and / or sigma change.

[0086] If one or more of the estimated mean and standard deviation values ​​are above the relevant threshold, then procedure 818 involves recursively estimating the air filter health status based on a change in the mean and / or standard deviation value. For example, the control may predict a higher deterioration state (the filter is more deteriorated) if one or more of the mean and standard deviation values ​​decrease. The change may include a change since the last routine run, a change since the last air filter maintenance, or a change since the filter was installed in the vehicle. As with the Fig. As described in sections 10-12, the mean MAF / MAP value measured at the filter can begin to decrease when the filter becomes clogged. Additionally, the data may be less scattered, resulting in a decrease in the standard deviation value. For example, the estimated air filter health status might include an estimate of the percentage of remaining filter life used. If the health status is 60%, for instance, this indicates that 60% of the filter's lifespan has expired, leaving only 40% of the air filter's remaining lifespan available for use.

[0087] Fig. Figure 9 shows an example of the effect of filter clogging on the statistical parameters of the measured mass airflow. Chart 900 depicts the MAF change at a throttle angle for a new (completely clean or unclogged) air filter at curve 902 (solid line), including a mean value and a standard deviation value 906 (solid line). Chart 900 also depicts the MAF change at a throttle angle for a completely clogged air filter at curve 904 (dashed line), including a mean value and a standard deviation value 908 (dashed line). In the example shown, at larger throttle angles, such as above 01, MAF varies between a value of α (MAF for new filter) and β (MAF for clogged filter).Based on the calibrated MAF mean and standard deviation value of the new filter compared to the clogged filter, the filter health can be determined using the following equation: %health=100∗(1α−βMAF−βα−β)

[0088] It will be understood that in alternative examples, MAP can be measured. In this case, MAP can vary between a value of α (MAP for a new filter) and β (MAP for a clogged filter) at larger throttle angles, such as above θ1. Based on the calibrated mean and standard deviation of the new filter compared to the clogged filter, the filter health can be determined using the following equation: %health=100∗(1α−βMAF−βα−β)

[0089] For a data set Xi, where i=1...N, and ordered data Xi = {x1, x2, ...xN},

[0090] For a dataset Xi, where i=1...N, and ordered data Xi = {x1, x2, ...xN}, assuming that for large N, µ(N+1) ~ µ(N); we obtain: σ(N)=∑i=1N(xi−μ(N))2N−1=σ(N+1)2=∑i=1N+1(xi−μ(N+1))2Nσ(N+1)2=∑i=1N+1(xi−μ(N+1))2+(xN+1−μ(N+1 ))2N−1(N−1N)σ(N+1)2∼∑i=1N(xi−μ(N))2+(xN+1−μ(N+1))2N−1(N−1N)∼[σ(N)2+(xN+1−μ(N+1))2N−1](N−1N) where: μ(N+1)=NN+1∗[μ(N)+xN+1N]:

[0091] The equations above show a step-by-step derivation of the recursive estimation of mu and sigma based on old values ​​(values ​​determined in the last iteration), new estimates (xN+1) and the current data counter (N).

[0092] With renewed reference to Fig. 8. The procedure at 820 involves updating the air filter health status based on vehicle driving statistics, including operator driving patterns and habits. The algorithm used to estimate the air filter health status may rely on a fixed deterioration rate. However, as with respect to Fig. As explained in section 2, events can occur that cause the deterioration rate to increase or decrease from the fixed rate. Some of these events can occur outside the vehicle's internal combustion engine, such as weather events, including precipitation (which can affect the component due to an increase in ambient humidity). Other events can involve operator driving patterns, such as the operator's tendency to drive aggressively when driving at a constant speed versus frequently using the accelerator and brake pedals, the average speed at which the vehicle is driven, the average mode in which the vehicle is driven (e.g., internal combustion engine or electric mode in a hybrid vehicle), etc.For example, if the operator frequently uses the accelerator and brake pedals or drives aggressively, the airflow through the filter may change more often, causing increased wear. Referring to the previous example, if the initial estimated health status based on driving statistics is 60%, the health status may be updated to 68%, indicating that the air filter is deteriorating more rapidly, or that this is to be expected, due to the vehicle's driving statistics.

[0093] In procedure 822, the process involves converting the estimated health status into an estimate of the remaining service life of the air filter, with the remaining service life estimate provided as a time or distance remaining before the air filter requires servicing. The controller can use the estimated health status and the estimated current vehicle distance (such as that indicated by the vehicle's odometer) to calculate the remaining service life. The conversion can be based on any of the vehicle's previous driving history data, including a previous history of airflow propagation, and predicted future driving. The conversion can further be based on a distance traveled by the vehicle, estimated by an odometer, and the presence of selected weather events.

[0094] With the 824, the control unit can display the estimated time / distance remaining until component degradation to the vehicle operator, for example, on a display in the vehicle's center console. It can display, for instance, that "the air filter needs to be replaced in 120 miles." This can provide the operator with a more understandable estimate of when the air filter will require maintenance. The control unit can also display the estimated remaining time or duration to a vehicle operator as a number of remaining fuel refueling events until the air filter needs maintenance. Furthermore, the displayed estimate can prompt the vehicle operator to adjust their driving pattern; for example, the operator might be encouraged to drive less aggressively.

[0095] For example, the control unit can indicate a deterioration state of an internal combustion engine intake air filter based on a spread of airflow readings that is smaller than expected when the throttle angle is above an upper threshold. The indication can also be based on an average of airflow readings that is smaller than expected when the throttle angle is above an upper threshold. The control unit can predict the deterioration state of the air filter based on the spread of airflow readings that is smaller than expected and convert the predicted deterioration state into an estimate of the remaining time or duration before the air filter requires maintenance, for display to a vehicle operator.The conversion can be based on any of the vehicle's previous driving history data, including a previous history of airflow measurement propagation, and predicted future driving behavior. The conversion can also be based on a distance traveled by the vehicle, estimated via an odometer, and the presence of selected weather events (such as blizzards, dust storms, and various precipitation types, such as snow and rain).

[0096] In this way, Fig. 8. An estimate of the remaining percentage of service life of an air filter is provided solely based on the commanded throttle angle and the measured mass airflow. Consequently, the method is able to provide an accurate estimate without requiring excessive memory or processor resources to assess the health of the air filter.

[0097] Exemplary data illustrating mu and sigma tendencies during air filter clogging are presented in the Fig. Shown 10-11. The map 1000 from Fig. Figure 10 presents MAF data collected at different commanded throttle angles for a new filter (thinner lines) and a 60% clogged filter (thicker lines). At larger throttle angles, such as above 55 degrees, the mean MAF for the clogged filter is lower. Additionally, the standard deviation, or dispersion, of the MAF data for the clogged filter is lower. In particular, several of the measured MAF data points are more closely grouped at larger throttle angles, whereas at the same throttle angles, there is significantly more variation for a clean filter. Map 1100 from Fig. Figure 11 shows a more detailed (enlarged) view of the data from Fig. 10 for throttle angles greater than 60 degrees. If only larger throttle angles are considered, the effect of other blockages (such as throttle clogging) on ​​the evaluation is reduced. The statistical parameters of the data for the new filter, as shown, include the mean (µ1) at the solid line 1152 and the standard deviation (σ1) at the solid arrow 1154. Similarly, the statistical parameters of the data for the 60% clogged filter, as shown, include the mean (µ2) at the dashed line 1156 and the standard deviation (σ2) at the dashed arrow 1158. At larger throttle angles, the mean MAF for the clogged filter is lower. Furthermore, the standard deviation or dispersion of the MAF data for the clogged filter is lower.In particular, several of the measured MAF values ​​are more closely grouped at larger throttle angles, whereas at the same throttle angles, there is significantly more variation for a clean filter. In this way, by monitoring the change in the statistical parameters for a measured MAF at larger throttle angles, a change in the condition of an intake filter can be predicted.

[0098] In this way, a control unit can predict a deterioration state of an internal combustion engine intake air filter based on a comparison of measured manifold airflow readings relative to commanded throttle angles during vehicle operation, after a threshold quantity of airflow readings has been collected above a threshold throttle angle, and then convert the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator based on previous driving history data and predicted future driving.Manifold airflow measurements can be collected during steady-state or transient combustion engine operating conditions via an airflow sensor coupled downstream of an intake throttle. Manifold airflow measurements collected during transient combustion engine operating conditions can be weighted more heavily than those collected during steady-state combustion engine operating conditions. Prediction based on comparison can involve deriving a standard deviation value and a mean value of the manifold airflow from each of the measured airflow values ​​and increasing the predicted deterioration state if one or more of the standard deviation value and the mean value fall below corresponding expected values.The derivation may place greater weight on manifold airflow measurements taken above the threshold throttle angle than on airflow measurements taken below the threshold throttle angle. Here, the corresponding expected values ​​are based on an initial standard deviation and an initial average value determined at the time the air filter was installed in the vehicle, and further, on the vehicle distance traveled since the air filter was installed. Alternatively, corresponding expected values ​​are based on previous driving history data, including a previous deterioration history of the air filter, and the corresponding expected values ​​include a most recent standard deviation and a most recent average value of the manifold airflow estimated during an immediately preceding repetition of the prediction. The data from the... Fig. 10-11 are in cards 1200 and 1250 Fig.Figure 12 is also shown graphically. As shown in chart 1200, the mean MAF value increases to a larger absolute value for a new filter compared to a clogged filter. By monitoring a change in the mean of the measured MAF, a change in the degree of filter clogging can be estimated, and the filter clogging rate over the duration or distance of vehicle operation can be estimated. This, in turn, can be used to predict the time remaining before the filter becomes 100% clogged. As shown in chart 1250, the MAF standard deviation value increases to a larger absolute value for a new filter compared to a clogged filter.By monitoring changes in the standard deviation of the measured MAF, it is possible to estimate changes in the filter's degree of clogging and to predict the filter's clogging rate over the duration or distance of vehicle operation. This, in turn, can be used to predict the time remaining before the filter becomes 100% clogged.

[0099] While the preceding examples demonstrate the correlating change in the statistical parameters of the measured MAF with air filter clogging, it will be understood that a change in the statistical parameters of the measured MAP can be similarly correlated with air filter clogging. At larger throttle angles, the effect of other blockages (such as throttle blockage) on the filter rating at a given intake pressure is reduced. When evaluating the air filter based on the MAP estimate, the measured MAP can be compared to the barometric pressure (BP). To reduce the effect of BP change on the estimated MAP measurements, a reference BP value can be selected. The variation of the current BP value relative to the reference value is subtracted from the mean MAP measurement.

[0100] This approach provides a predictive method for assessing the remaining service life of a vehicle component. The predictive method can be used to complement existing predictive features that primarily approximate a component's remaining service life as a ratio (a percentage) by estimating the remaining time and / or distance until the component deteriorates. By relying on the statistical characterization of the component's health as a function of time and distance traveled, the remaining service life can be determined more accurately and presented to the vehicle operator as a more understandable metric.

[0101] An exemplary procedure for a vehicle comprises: predicting the deterioration state of a vehicle component based on a specified metric derived from a captured vehicle operating parameter, including a past history of the specified metric; and converting the predicted deterioration state into a remaining time or duration for display to a vehicle operator based on past driving history data and predicted future driving, including the past history of the specified metric. In the preceding example, the vehicle component is additionally or optionally a system battery, the specified metric is one or more of a battery resistance and a battery capacity, and the captured vehicle operating parameter includes one or more of a battery current and a battery voltage.In one or all of the preceding examples, the prediction additionally or optionally includes predictions of a higher deterioration state with increasing battery resistance or decreasing battery capacity. In one or all of the preceding examples, the vehicle component additionally or optionally includes an internal combustion engine intake air filter, wherein the determined metric is one or more of a mean and a standard deviation value of airflow through the filter, and the detected vehicle operating parameter includes manifold airflow. In one or all of the preceding examples, the method additionally or optionally further includes giving greater weight to the manifold airflow detected at higher intake throttle angles than to the manifold airflow detected at lower throttle angles.In one or all of the preceding examples, the recorded vehicle operating parameter additionally or optionally includes manifold pressure relative to atmospheric pressure. In one or all of the preceding examples, the prediction additionally or optionally includes predictions of a higher deterioration state when one or more of the mean and standard deviation values ​​decrease. In one or all of the preceding examples, the recorded vehicle operating parameter is additionally or optionally recorded during transient and steady-state vehicle operating conditions, and the procedure further includes weighting the recorded vehicle operating parameter recorded during transient vehicle operating conditions more heavily than the recorded vehicle operating parameter recorded during steady-state vehicle operating conditions.In one or all of the preceding examples, the procedure additionally or optionally includes estimating a value of the specified metric as a function of a recent estimate of the specified metric, retrieved from the previous history of the specified metric, and a distance traveled by the vehicle since the most recent estimate of the specified metric. In one or all of the preceding examples, the procedure additionally or optionally includes estimating a value of the specified metric as a function of an initial estimate of the specified metric, retrieved from the previous history of the specified metric at a point in time when the component was installed in the vehicle. In one or all of the preceding examples, the conversion is additionally or optionally based on an absolute distance traveled by the vehicle, retrieved from previous driving history data.In one or all of the preceding examples, the method additionally or optionally includes converting the predicted deterioration state into a remaining number of fuel refueling events for display to the vehicle operator based on previous driving history data and predicted future driving.

[0102] Another exemplary procedure for a vehicle includes: predicting the deterioration state of a vehicle component based on a change in a metric associated with the vehicle component over a period of time, the metric being derived from a captured vehicle operating parameter; and converting the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator based on past driving history data and predicted future driving. In the preceding example, the procedure additionally or optionally further includes displaying the estimate of the remaining time or duration to the vehicle operator as a number of refueling events remaining until the vehicle component requires service.In one or all of the preceding examples, the past driving pattern and the predicted future driving pattern additionally or optionally include one or more measures of driving aggressiveness, a rate of pedal use, frequent driving time patterns, habitual probability patterns, route-based statistical profiles, and environmental attribute profiles, where the environmental attribute profiles include past and predicted weather events along a vehicle's route. In one or all of the preceding examples, the conversion based on distance traveled additionally or optionally includes conversion by a factor based on a current vehicle odometer reading.In one or all of the preceding examples, the vehicle component is additionally or optionally a system battery, the detected vehicle operating parameter includes a battery current and a battery voltage, and where, if the vehicle component is an internal combustion engine intake air filter, the detected vehicle operating parameter includes a manifold airflow and a manifold air pressure.In one or all of the preceding examples, the change in the metric additionally or optionally involves a change from an initial value of the metric estimated at an installation time of the vehicle component, and wherein the prediction involves increasing the deterioration state of the vehicle component towards a fully deteriorated state when a difference between a current value of the metric relative to the initial value of the metric exceeds a threshold, the threshold being based on a distance traveled by the vehicle over the duration.

[0103] Another exemplary procedure for a vehicle system comprises: predicting a deterioration state of a vehicle battery based on a change in the battery resistance and capacity from initial values ​​at an installation time in the vehicle system over a vehicle driving duration and further based on a distance traveled by the vehicle over the duration, wherein the battery resistance and capacity are derived from a detected battery current and a detected battery voltage; and converting the predicted deterioration state into an estimate of the remaining time or duration before the battery requires servicing, for display to a vehicle operator, wherein the conversion is based on any of the previous driving history data and predicted future driving.In the preceding example, the battery voltage or current recorded during transient vehicle operating conditions is additionally or optionally weighted differently from the battery voltage or current recorded during stationary vehicle operating conditions.

[0104] Another exemplary procedure for a vehicle includes: indicating a deterioration state of an internal combustion engine intake air filter based on a spread of airflow measurements that is smaller than expected when the throttle angle is above an upper threshold. In the preceding example, the indication is additionally or optionally further based on an average of the airflow measurements that is smaller than expected when the throttle angle is above an upper threshold. In one or all of the preceding examples, the indication additionally or optionally includes: predicting the deterioration state of the air filter based on the spread of the airflow measurements that is smaller than expected; and converting the predicted deterioration state into an estimate of the remaining time or duration before the air filter requires servicing, for display to a vehicle operator.In one or all of the preceding examples, the conversion is additionally or optionally based on any of the vehicle's previous driving history data, including a previous history of airflow measurement propagation, and predicted future driving behavior of the vehicle. In one or all of the preceding examples, the conversion is additionally or optionally further based on a distance traveled by the vehicle, estimated by an odometer, and the presence of selected weather events. In one or all of the preceding examples, the prediction additionally or optionally takes place after a threshold amount of data has been collected, with an intake throttle at the throttle angle above the upper threshold.In one or all of the preceding examples, the method additionally or optionally includes displaying to a vehicle operator an estimate of the remaining time or duration as a number of remaining fuel refueling events until the air filter requires service. In one or all of the preceding examples, the expected propagation of the airflow readings is additionally or optionally based on each of the distance traveled by the vehicle, previous vehicle driving history data, predicted future vehicle driving, and an initial propagation of the airflow readings at the time the filter is installed in the vehicle.In one or all of the preceding examples, the expected average of the airflow measurements is additionally or optionally based on each of the distance traveled by the vehicle, previous driving history data of the vehicle, predicted future driving of the vehicle, and an initial average of the airflow measurements at a time when the filter was installed in the vehicle.

[0105] Another exemplary procedure for a vehicle includes: predicting a deterioration state of an internal combustion engine intake air filter based on a comparison of measured manifold airflow readings relative to commanded throttle angles during vehicle operation, after a threshold quantity of airflow readings has been collected above a threshold throttle angle; and converting the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator based on past driving history data and predicted future driving.In the preceding example, the manifold airflow measurements are additionally or optionally collected during steady-state or transient combustion engine operating conditions via an airflow sensor coupled downstream of an intake throttle, and the manifold airflow measurements collected during transient combustion engine operating conditions are weighted more heavily than the manifold airflow measurements collected during steady-state combustion engine operating conditions.In one or all of the preceding examples, prediction based on comparison additionally or optionally includes: deriving each of a standard deviation value and a mean value of the manifold airflow based on the measured airflow readings; and increasing the predicted deterioration state if one or more of the standard deviation value and the mean value fall below corresponding expected values. In one or all of the preceding examples, derivation additionally or optionally includes giving greater weight to the manifold airflow readings measured above the threshold throttle angle than to the airflow readings measured below the threshold throttle angle.In one or all of the preceding examples, derivation additionally or optionally involves giving greater weight to the manifold airflow measured at throttle angles above a threshold than to the manifold airflow measured at throttle angles below a threshold. In one or all of the preceding examples, the corresponding expected values ​​are additionally or optionally based on an initial standard deviation value and an initial mean value determined at a time when the air filter was installed in the vehicle, and furthermore on the basis of a vehicle distance traveled since the air filter was installed in the vehicle.In one or all of the preceding examples, the method additionally or optionally further comprises that the relevant expected values ​​are based on previous driving history data, including a previous deterioration history of the air filter, and wherein the relevant expected values ​​include a recent standard deviation value and a recent average value of the manifold airflow, which are estimated during an immediately preceding repetition of the prediction.

[0106] An exemplary vehicle system comprises: an internal combustion engine, including an intake manifold; an air filter coupled to the intake manifold; an intake throttle; a manifold airflow sensor coupled downstream of the intake throttle; and a control system.The controller is configured with computer-readable instructions stored in non-volatile memory for the following: storing the measured airflow readings when the intake throttle is commanded to exceed a threshold throttle angle; estimating a metric that specifies a distribution of manifold airflow based on the stored measured airflow readings; predicting an air filter deterioration state based on the estimated metric relative to a threshold; and converting the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator based on past driving history data and predicted future driving, including a past history of the estimated metric. In the preceding example, predicting involves predicting a higher deterioration state if the estimated metric falls below the threshold.In one or all of the preceding examples, the metric is additionally or optionally a first metric, and the control further includes instructions for estimating a second metric that specifies an average manifold airflow through the air filter, and wherein the predicting includes predictions of the higher deterioration state when the second metric falls below the threshold. In one or all of the preceding examples, the threshold is additionally or optionally determined as a function of a most recent estimate of the metric, retrieved from the earlier history of the estimated metric, and a distance traveled by the vehicle since the most recent estimate of the estimated metric.In one or all of the preceding examples, the threshold is additionally or optionally determined as a function of an initial estimate of the metric at an air filter installation time, retrieved from the previous course of the estimated metric, and a distance traveled by the vehicle since the air filter was installed.

[0107] In another representation, the procedure can involve comparing a statistically determined metric of the recorded vehicle operating parameter with an initial statistical attribute of the recorded vehicle operating parameter, which was recorded at an installation time of the vehicle component, and increasing the predicted deterioration state if the current statistical attribute falls relative to the initial statistical attribute.In another representation, predicting the deterioration state of an air filter may involve retrieving a first mean and first standard deviation value of the detected airflow at a point in time when the component was installed in the vehicle; comparing the first mean with a second mean of the detected airflow at a current point in time; comparing the first standard deviation value with a second standard deviation value of the detected airflow at the current point in time; and increasing the predicted deterioration state of the air filter towards a fully deteriorated state if the second mean falls below the first mean or the second standard deviation value falls below the first standard deviation value.In the preceding example, the procedure may additionally or optionally further include, in response to the second mean or second standard deviation value falling below a threshold, indicating that the air filter is completely deteriorated and must be replaced. In the preceding example, the manifold airflow is additionally or optionally recorded during steady-state and transient operating conditions, and the prediction involves giving greater weight to the vehicle operating parameters recorded during transient operating conditions than to those recorded during steady-state operating conditions.

[0108] It should be noted that the exemplary control and estimation routines included herein can be used with various internal combustion engine and / or vehicle system configurations. The control methods and sequences disclosed herein can be stored as executable instructions in non-volatile memory and executed by the control system, including the controller, in combination with the various sensors, actuators, and other internal combustion engine hardware. The specific routines described herein can represent one or more from any number of processing strategies, such as event-driven, interrupt-driven, multitasking, multithreading, and the like. Accordingly, various illustrated actions, processes, and / or functions can be performed in the illustrated sequence or in parallel, or in some cases, omitted.Similarly, the processing sequence is not strictly necessary to achieve the features and advantages of the exemplary embodiments described here, but is provided to facilitate illustration and description. One or more of the illustrated actions, processes, and / or functions can be performed repeatedly, depending on the specific strategy employed. Furthermore, the described actions, processes, and / or functions can graphically represent code to be programmed into non-volatile memory of the computer-readable storage medium in the internal combustion engine control system, with the described actions being executed by carrying out the instructions in a system that includes the various internal combustion engine hardware components in combination with the electronic control unit.

[0109] It is understood that the interpretations and routines disclosed herein are exemplary in nature and that these specific embodiments are not to be interpreted in a restrictive sense, as numerous variations are possible. For example, the aforementioned technology can be applied to V-6, I-4, I-6, V-12, 4-cylinder boxer, and other types of internal combustion engines. The subject matter of this disclosure includes all novel and non-obvious combinations and sub-combinations of the different systems and configurations, and other features, functions, and / or properties disclosed herein.

Claims

[1] Methods for a vehicle, comprising: Recursive prediction of a deterioration state of a vehicle component by updating a previously detected deterioration state of the vehicle component, wherein the update is based on a vehicle operating parameter detected by a sensor, wherein the previously detected deterioration state is based on a specific metric, including a previous history of the specific metric; and Converting the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator on a display, wherein the conversion is based on previous driving history data and on predicted future driving, including the previous history of the specified metric; wherein the vehicle component is an intake air filter of the internal combustion engine and the determined metric is one or more of a mean and a standard deviation value of airflow through the filter and the detected vehicle operating parameter includes manifold airflow. [2] Method according to claim 1, further comprising giving greater weight to the manifold airflow captured at higher intake throttle angles than to the manifold airflow captured at lower throttle angles. [3] Method according to claim 1, wherein the recorded vehicle operating parameter further includes manifold air pressure relative to atmospheric pressure. [4] Method according to claim 1, wherein the prediction includes predictions of a higher deterioration state when one or more of the mean and standard deviation values ​​decrease. [5] Method according to claim 1, wherein the recorded vehicle operating parameter is recorded during temporary and stationary vehicle operating conditions, the method further comprising weighting the recorded vehicle operating parameter recorded during temporary vehicle operating conditions more heavily than the recorded vehicle operating parameter recorded during stationary vehicle operating conditions. [6] Method according to claim 1, further comprising estimating a value of the specified metric as a function of a most recent estimate of the specified metric, retrieved from the previous course of the specified metric, and a distance traveled by the vehicle since the most recent estimate of the specified metric. [7] Method according to claim 1, further comprising estimating a value of the determined metric as a function of an initial estimate of the determined metric, retrieved from the previous course of the determined metric at an installation time of the vehicle component in the vehicle. [8] Method according to claim 1, wherein the conversion is further based on an absolute distance traveled by the vehicle, retrieved from the previous driving history data. [9] Method according to claim 1, further comprising converting the predicted deterioration state into a remaining number of fuel refueling events for display to the vehicle operator based on previous driving history data and predicted future driving. [10] Methods for a vehicle, comprising: Recursive prediction of a deterioration state of a vehicle component by updating a previously detected state of deterioration of the vehicle component, wherein the update is based on a vehicle operating parameter detected by a sensor, a change in a metric associated with the vehicle component over a duration, and a distance traveled by the vehicle over that duration; Converting the predicted deterioration state into an estimate of the remaining time or duration for display to a vehicle operator on a display, wherein the conversion is based on previous driving history data and predicted future driving; and Displaying the estimate of remaining time or duration to the vehicle operator as a number of refueling events remaining until the vehicle component requires service, wherein the past driving history and predicted future driving include one or more of a measure of driving aggressiveness, a rate of pedal usage, frequent driving time patterns, habitual probability patterns, route-based statistical profiles, and environmental attribute profiles, wherein the environmental attribute profiles include past and predicted weather events along a vehicle driving route, and wherein the conversion based on past driving history and predicted future driving includes conversion by a factor based on a current vehicle odometer reading.

Images