FUEL SYSTEM DIAGNOSIS

Abstract

A method for a fuel system (18) for a vehicle, comprising: during a fuel system leak test and in response to an unintentional, temporary closing of a mechanical valve (106A, 106B, 108) coupled to a fuel tank (20), not terminating the fuel system leak test, characterized in that the fuel system leak test comprises: closing a tank vent valve (114) coupled between a fuel vapor reservoir (22) and the atmosphere; opening a purge valve (112) coupled between a fuel vapor reservoir (22) and an engine inlet; applying an engine inlet vacuum to the fuel tank (20) with the mechanical valve (106A, 106B, 108) open;and after installation, isolating the fuel tank (20) by closing the purge valve (112) while the mechanical valve (106A, 106B, 108) is still open, monitoring a vacuum drain in the fuel tank (20) and indicating a fuel system leak based on a vacuum drain rate.;

Description

The present description concerns systems and methods for improving the accuracy of fuel system leak detection in a vehicle, such as a hybrid vehicle. Vehicles can be equipped with fuel evaporation systems to reduce the release of fuel vapors into the atmosphere. For example, evaporated hydrocarbons (HCs) from a fuel tank can be stored in a fuel vapor recovery tank filled with an adsorbent that adsorbs and stores the vapors. Later, when the engine is running, the fuel evaporation system allows the vapors to be purged into the engine intake manifold for use as fuel. Since leaks in the exhaust aftertreatment system can unintentionally allow fuel vapors to escape into the atmosphere, leak detection routines can be performed intermittently when the engine is not running. During this process, the system is sealed after applying a vacuum to the fuel system, and the pressure drop rate is monitored. By comparing the actual pressure drop with a reference value (such as one determined by a reference orifice), leaks can be identified. To prevent false-positive leak detection, vehicle control systems can abort or delay leak tests when predefined conditions are met. An exemplary approach to reducing false-positive leak detection is shown by Suzuki in US 6,973,924 B1. When refueling a fuel tank is detected, a leak test routine is delayed until a tank purge threshold has been reached. In particular, a leak test is not performed under conditions where refueling generates a large amount of evaporated fuel, as the refueling vapors can increase the possibility of a false-positive leak detection. However, the present inventors have identified potential problems with such an approach. For example, Suzuki's approach may not adequately address false leak detections resulting from the unintentional temporary closing (also known as clogging) of one or more mechanical fuel tank vent valves. Specifically, leak diagnosis may be performed with the engine running while the vehicle is in motion. During this process, dynamic vehicle maneuvers, such as spirited cornering, climbing a hill, or driving along a bumpy road, can cause fuel to slosh around and temporarily clog one or more passive tank vent valves (which would otherwise be expected to be open during leak diagnosis).Should this occur, the fuel tank can be isolated, and the volume of the exhaust aftertreatment system is dramatically reduced. If a leak test is performed when an unintentional valve closure occurs, false leak detection may result because leak detection reference pressure values ​​are based on a fuel tank's fill volume. If a fuel tank is isolated due to an unintentional, temporary closure of a fuel tank vent valve, the likelihood of a false leak detection increases. This reduces the reliability of the leak test. This can lead to the triggering of an unwarranted fault indication, possibly accompanied by a warning light, and thus to unnecessary maintenance. EP 0 733 793 A2 discloses an evaporation emission control system that prevents the emission of fuel vapor into the atmosphere and is capable of detecting the occurrence of a fault in the system. The technical problem to be solved can be seen as eliminating or at least reducing the disadvantages of the prior art. According to the invention, this problem is solved by the subject matter of the independent claims. The above problems are solved by a method for a vehicle fuel system according to the subject matter of claim 1. The method comprises: during a fuel system leak test and in response to an unintentional, temporary closing of a mechanical valve coupled to a fuel tank, not terminating the fuel system leak test. Instead, a leak test is repeated, thus reducing false leak detections. As an example, an engine fuel system leak test can be initiated by opening a purge valve. It is therefore expected that one or more passive, mechanical vent valves connected to the fuel tank will be open during the leak test. An engine intake vacuum can then be applied to the fuel system. While the vacuum in the fuel tank is increased, the fuel tank pressure can be monitored. A sudden fluctuation in the fuel tank pressure experienced during the (initial) increase in vacuum may indicate an unintentional, temporary closing (also referred to here as clogging) and subsequent opening (also referred to here as releasing) of a fuel tank vent valve.For example, the vacuum may suddenly increase faster than expected, suggesting an unintentional closure of a vent valve, followed by a sudden decrease back to the expected profile, suggesting the vent valve reopening. In one example, the leak test may be performed while a vehicle is in motion, and the momentary closure of the vent valve could be caused by certain vehicle maneuvers (for example, a sharp turn). In response to the indication of an unintentional, temporary closure of the vent valve, the fuel system leak test can be interrupted, not terminated. Instead, the fuel tank vacuum can be released, the purge valve closed, and the fuel tank settings from before the leak test resumed. Once the fuel tank pressure has stabilized, the fuel system leak test can be restarted. Specifically, the purge valve can be reopened, and the vacuum in the fuel tank can be increased again. If no pressure fluctuation is present during the (second or subsequent) vacuum increase, it can be determined that valve clogging did not occur this time. Accordingly, the fuel tank can be isolated after the most recent vacuum application (by closing the purge valve), and the release of the vacuum to atmospheric pressure can be monitored.A fuel system leak can then be identified based on the vacuum release rate. If, for example, the vacuum release is faster than a threshold rate, then a fuel system leak is confirmed. In other embodiments, an unintentional, temporary closure of the fuel tank vent valve may be detected due to a pressure fluctuation experienced during vacuum release. For example, the vacuum may be released more quickly than expected, suggesting an unintentional closure of the vent valve, followed by a sudden return to the expected pressure profile, suggesting the vent valve reopening. If the indication is received during vacuum release, the fuel system leak test may be interrupted, not completed. That is, the vacuum release data is disregarded while initial fuel system settings (those prior to initiating the leak test) are resumed. Once the fuel tank pressure has stabilized, the fuel system leak test can then be restarted. In particular, the vacuum in the fuel tank can be increased again.If no pressure fluctuation is present during the vacuum increase and subsequent vacuum release, it can be determined that valve clogging was not present this time. Accordingly, the most recent vacuum release data can be used to identify a fuel system leak. In this way, false leak detections can be reduced by aborting a fuel system leak test if an unintentional, momentary closure of a fuel tank vent valve is detected. By resuming initial fuel system settings from before the test and re-running the leak test after fuel tank pressures have stabilized following the aborted test, leak tests can be completed with more reliable results. By relying solely on vacuum discharge data from a leak test when no valve blockage has been determined, fuel system leaks can be identified accurately and reliably. Fig. 1 shows a schematic representation of a vehicle fuel system. Fig. 2 shows a detailed flowchart representing a routine that can be implemented to perform a fuel system leak test.Figure 3 shows a detailed flowchart representing a routine that can be implemented to identify an unintentional, transient opening of a fuel tank vent valve during the leak test of Figure 2. Figure 4 shows an expected fuel tank pressure profile during a vacuum intensification phase and a vacuum drain phase of a fuel system leak test. Figures 5-11 show deviations in the fuel tank pressure profile during the vacuum intensification and / or vacuum drain phase of a fuel system leak test due to a brief, unintentional opening and subsequent closing of a fuel tank vent valve. Figure 12 shows an example of a fuel system leak test with an unintentional, transient opening of a fuel tank vent valve during the leak test. Methods and systems for identifying leaks in a fuel system coupled to a vehicle engine, such as the fuel system shown in Fig. 1, are provided. A vacuum leak test can be performed on the fuel system with the engine running while the vehicle is in motion. A controller can be configured to execute a control routine, such as the exemplary routine shown in Fig. 2, to apply an engine inlet vacuum to the fuel system and determine a fuel system leak based on the rate of subsequent vacuum release. The controller can also execute a routine, such as the one shown in Fig. 3, to identify a temporary unintentional closure of a fuel tank vent valve based on fuel tank pressure fluctuations experienced during a vacuum intensification or vacuum release phase of the leak test.The control system can only terminate the leak test if no pressure fluctuations are detected during the test. Otherwise, if a temporary, unintentional closure of a fuel tank vent valve is detected during the leak test, the control system may interrupt the test and repeat it at a later time. Exemplary pressure deviations and fluctuations resulting from a momentary blockage of the tank vent valve are shown with reference to Figures 5-11 and compared with an expected leak test pressure profile shown in Figure 4. Figure 12 describes an exemplary leak test procedure. In this way, false leak detections can be reduced, and the reliability of a fuel system leak test can be improved. Fig. 1 shows a schematic representation of a hybrid vehicle system 6, which can derive drive power from the engine system 8 and / or an on-board energy storage device (not shown), such as a battery system. An energy conversion device, such as a generator (not shown), can be operated to absorb energy from the vehicle's motion and / or the operation of the internal combustion engine and then convert the absorbed energy into a form of energy suitable for storage by the energy storage device. The engine system 8 can include a multi-cylinder engine 10 30. The engine 10 includes an engine inlet 23 and an engine outlet 25. The engine inlet 23 includes an air inlet throttle 62, which is flow-coupled to the engine inlet manifold 44 via an inlet channel 42. Air can enter the inlet channel 42 via an air filter 52. The engine outlet 25 includes an exhaust manifold 48, which leads to an exhaust port 35 that discharges exhaust gas to the atmosphere. The engine outlet 25 can include one or more exhaust aftertreatment devices 70, which are mounted in a closely coupled position. The one or more exhaust aftertreatment devices can include a three-way catalytic converter, a lean NOx trap, a diesel particulate filter, an oxidation catalyst, etc. It is understood that the engine can include other components, such as various valves and sensors, as further described below.In some embodiments where the engine system 8 is a turbocharged engine system, the engine system may further include a charging device, such as a turbocharger (not shown). The engine system 8 is coupled to a fuel system 18. The fuel system 18 includes a fuel tank 20, which is coupled to a fuel pump 21 and a fuel vapor reservoir 22. The fuel tank 20 receives fuel via a refueling line 116, which acts as a passage between the fuel tank 20 and a refueling hatch 129 on an external body panel of the vehicle. During a refueling event, fuel can be pumped into the vehicle from an external source through a refueling inlet 107. During a refueling event, one or more fuel tank vent valves 106A, 106B, 108 (described in more detail below) can be open to allow refueling vapors to be directed to the reservoir 22 and then stored. The fuel tank 20 can hold several fuel mixtures, including fuels with different alcohol concentrations, such as various gasoline-ethanol blends, including E10, E85, gasoline, etc., and combinations thereof. A fuel level sensor 106 positioned in the fuel tank 20 can indicate the fuel level to the controller 12 (“fuel level input”). As shown, the fuel level sensor 106 can include a float connected to a control resistor. Alternatively, other types of fuel level sensors can also be used. The fuel pump 21 is configured to pressurize fuel, which is then delivered to the injectors of the engine 10, such as the exemplary injector 66. Although only a single injector 66 is shown, additional injectors are provided for each cylinder. It is understood that the fuel system 18 can be a non-return fuel system, a return fuel system, or various other types of fuel systems. Vapors generated in the fuel tank 20 can be routed via line 31 to a fuel vapor reservoir 22 before being purged to the engine inlet 23. The fuel tank 20 can contain one or more vent valves for venting vapors generated throughout the day and refueling vapors produced in the fuel tank to the fuel vapor reservoir 22. The one or more vent valves can be electronically or mechanically actuated and can be active vent valves (i.e., valves with moving parts that are actuated to the open or closed position by a control mechanism) or passive valves (i.e., valves without moving parts that are actuated to the open or closed position based on the tank level).In the example shown, the fuel tank contains 20 gas vent valves (GVV - gas vent valve) 106A, 106B at both ends of the fuel tank 20 and one fuel level vent valve (FLVV - fuel level vent valve) 108, all of which are passive vent valves. Each of the vent valves 106A, 106B, 108 may contain a tube (not shown) that extends to a varying degree into a vapor space 104 of the fuel tank. Based on a fuel level 102 relative to the vapor space 104 in the fuel tank, the vent valves may be open or closed. For example, the GVV 106A, 106B may extend less deeply into the vapor space 104, so they are normally open. This allows vapors and lost vapors generated during the day to be discharged from the fuel tank into container 22, thus preventing excessive pressure build-up in the fuel tank.When the vehicle is operating on an incline, if the fuel level 102 is artificially raised on at least one side of the fuel tank, the vent valve 106A, 106B can close to prevent liquid fuel from entering the vapor line 31. Alternatively, the FLVV 108 can extend further into the vapor space 104, so that it is normally open. This prevents overfilling of the fuel tank. Specifically, when refueling with a raised fuel level 102, the vent valve 108 can close, causing a pressure buildup in the vapor line 109 (located downstream of the refueling inlet 107 and connected to the channel 31) and at a filler neck connected to the fuel pump. The pressure increase at the filler neck can then activate the fuel pump, automatically stopping the fuel filling process and preventing overfilling. One problem with passive fuel tank vent valves is that during certain vehicle maneuvers, such as spirited cornering, hill climbing, or driving along a bumpy road, fuel can slosh around and cause the valve, which would otherwise be expected to be open, to close briefly and unintentionally. Other maneuvers can also cause the valve to reopen. If a vent valve is temporarily blocked, the fuel tank can become isolated, dramatically reducing the fuel system's volume. If an inadvertent closure of a fuel tank vent valve occurs during a fuel system leak test (described below), leak test data can be corrupted and false diagnostic codes can be triggered. As shown below with reference to Fig.2-3 implemented, engine control systems can be configured to identify a vent valve blockage during a leak test based on deviations in fuel tank pressure profiles during the leak test. In response to the identification of a closed vent valve when it should have been open, the leak test is aborted and repeated. This reduces the likelihood of false leak detection and improves vehicle fuel system warranties. It is understood that while the illustrated embodiment shows the ventilation valves 106A, 106B, and 108 as passive valves, in alternative embodiments one or more of them may be configured as electronic valves that are electronically coupled (via wiring) to a controller. In this case, a controller can send a signal to actuate the ventilation valves to the open or closed position. Furthermore, the valves may include electronic feedback to communicate an open / closed status to the controller. While the use of electronic ventilation valves with electronic feedback can allow a controller to directly determine whether a ventilation valve is open or closed (for example, to determine whether a valve is closed when it should have been open), such electronic valves can significantly increase the cost of such a fuel system.Furthermore, the wiring required to connect such electronic vent valves to the control system can act as a potential ignition source within the fuel tank, thus increasing the risk of fire in the fuel system. Therefore, by using passive fuel tank vent valves and monitoring fuel tank pressures during a leak test, vent valve blockages can be reliably identified without increasing the risk of fire in fuel systems. Referring again to Fig. 1, a fuel vapor reservoir 22 is filled with a suitable adsorbent for temporarily capturing fuel vapors (including vaporized hydrocarbons) generated during fuel refueling operations, as well as vapors produced throughout the day. In one example, the adsorbent used is activated carbon. When purging conditions are met, for example, when the reservoir is saturated, vapors stored in the fuel vapor reservoir 22 can be purged by opening the reservoir purge valve 112 via the purge line 28 to the engine inlet 23. Although a single reservoir 22 is shown, it is understood that the fuel system 18 can contain any number of reservoirs. The container 22 includes a vent 27 for directing gases from the container 22 to the atmosphere when storing or collecting fuel vapors from the fuel tank 20. The vent 27 can also allow fresh air to be drawn into the fuel vapor container 22 via the purge line 28 and the purge valve 112 to the engine inlet 23 when purging stored fuel vapors. Although this example shows that the vent 27 communicates with fresh, unheated air, various modifications are possible. The vent 27 can include a container vent valve 114 to regulate the airflow and vapors between the container 22 and the atmosphere. Furthermore, the container vent valve can be used for diagnostic routines.If present, the vent valve can be opened during fuel vapor storage operations (for example, when refueling and the engine is not running) so that air, without the fuel vapor after passing through the reservoir, can be expelled to the atmosphere. Similarly, the vent valve can be open during purging operations (for example, during reservoir regeneration and while the engine is running) to allow a flow of fresh air to remove the fuel vapors stored in the reservoir. Accordingly, the hybrid vehicle system 6 can exhibit reduced engine operating times because the vehicle is powered by an engine system 8 under some conditions and by the energy storage device under other conditions. Although the reduced engine operating times reduce the overall carbon emissions from the vehicle, they can also lead to insufficient purging of fuel vapors from the vehicle's evaporative cooling system. To counteract this, a fuel tank shut-off valve (not shown) can optionally be included in line 31, so that the fuel tank 20 is connected to the reservoir 22 via the shut-off valve. If included, the shut-off valve can be kept closed during engine operation to limit the amount of vapors generated throughout the day that are routed from the reservoir 22 to the fuel tank 20.During refueling operations and under selected purging conditions, the shut-off valve can be temporarily opened to direct fuel vapors from fuel tank 20 to reservoir 22. By opening the valve under purging conditions, when the fuel tank pressure is higher than a threshold (for example, above a limit of the mechanical pressure of the fuel tank beyond which the fuel tank and other fuel system components could be mechanically damaged), the refueling vapors can be released into the reservoir, and the fuel tank pressure can be maintained within pressure limits. One or more pressure sensors 120 can be coupled to a fuel system 18 to provide an estimate of the fuel system pressure. In one example, the fuel system pressure is a fuel tank pressure, where the pressure sensor 120 is a fuel tank pressure sensor coupled to the fuel tank 20 to estimate a fuel tank pressure or vacuum level. Although in the example shown the pressure sensor 120 is coupled between the fuel tank and the container 22, in alternative embodiments the pressure sensor can be coupled directly to the fuel tank 20. Fuel vapors released from reservoir 22, for example during a purging process, can be routed to the engine intake manifold 44 via purging line 28. The flow of vapors along purging line 28 can be regulated by a reservoir purge valve 112, which is coupled between the fuel vapor reservoir and the engine intake. The quantity and rate of vapors released by the reservoir purge valve can be determined by the duty cycle of an associated reservoir purge valve solenoid (not shown). Accordingly, the duty cycle of the reservoir purge valve solenoid can be determined by the vehicle's powertrain control module (PCM), such as control unit 12, which responds to engine operating conditions, including, for example, engine speed / load conditions, air-fuel ratio, reservoir load, etc.By controlling the tank purge valve to the closed position, the controller can seal the fuel vapor recovery system against the engine intake. An optional tank check valve (not shown) may be included in purge line 28 to prevent intake manifold pressure from forcing gases in the opposite direction to the purge flow. Thus, the check valve may be necessary if the tank purge valve control is not precisely timed, or if the tank purge valve itself may be forced open by high intake manifold pressure. An estimate of the manifold absolute pressure (MAP) can be obtained from the MAP sensor 118, which is coupled to the intake manifold 44 and connected to the controller 12.Alternatively, MAP can be derived from other engine operating conditions, such as mass air flow (MAF), as measured by a MAF sensor coupled to the intake manifold (not shown). The fuel system 18 can be operated in several modes by the controller 12 through targeted adjustment of the various valves and solenoids. For example, the fuel system can be operated in a fuel vapor storage mode, in which the controller 12 closes the canister purge valve (CPV) 112 and opens the canister vent valve 114 to direct refueling vapors and vapors generated during the day into the tank 22, while preventing fuel vapors from entering the intake manifold. As another example, the fuel system can be operated in a refueling mode (for example, when a driver requests refueling), in which the controller 12 can keep the canister purge valve 112 closed to depressurize the fuel tank before allowing fuel to be added.Thus, during both the fuel storage and refueling modes, it is assumed that the fuel tank vent valves 106A, 106B and 108 are open. As a further example, the fuel system can be operated in a tank purging mode (for example, after an exhaust aftertreatment system activation temperature has been reached and with the engine running), whereby the controller 12 can open the tank purging valve 112 and the tank venting valve 11. Thus, it can be assumed that during tank purging, the fuel tank venting valves 106A, 106B, and 108 are open (although in some embodiments, some valve combinations may be closed). In this mode, a vacuum generated by the intake manifold of the running engine can be used to draw fresh air through the vent 27 and through the fuel vapor reservoir 22 to draw the stored fuel vapors into the intake manifold 44. In this mode, the purged fuel vapors from the reservoir are combusted in the engine.Purging can continue until the amount of stored fuel vapor in the reservoir falls below a threshold. During purging, the learned vapor quantity / concentration can be used to determine the amount of fuel vapor stored in the reservoir, and then, in a later stage of the purging process (when the reservoir is sufficiently purged or empty), the learned vapor quantity / concentration can be used to estimate the state of charge of the fuel vapor reservoir. For example, one or more oxygen sensors (not shown) can be coupled to reservoir 22 (for example, downstream of the reservoir) or positioned in the engine inlet and / or outlet to provide an estimate of a reservoir load (i.e., an amount of fuel vapor stored in the reservoir).Based on the tank load and further on the basis of engine operating conditions, such as engine speed / load conditions, a flushing flow rate can be determined. The control unit 12 can also be configured to intermittently perform leak detection routines on the fuel system 18 to confirm that the fuel system is not compromised. Thus, leak detection routines can be performed while the vehicle is running, with the engine on (for example, during an engine mode of hybrid vehicle operation) or with the engine off (for example, during a battery mode of hybrid vehicle operation). Leak tests performed with the engine off can include applying a natural vacuum to the fuel system with the engine off. This can be achieved by sealing the fuel tank when the engine is switched off by closing the tank purge valve and the tank vent valve. As the fuel tank cools, a vacuum is created in the vapor space of the fuel tank (due to the relationship between temperature and pressure of gases).The tank vent valve is then opened, and the vacuum drop rate from the fuel tank is monitored. If the fuel tank pressure stabilizes at atmospheric pressure faster than expected, a fuel system leak is identified. Leak tests performed with the engine running involve applying an engine intake vacuum to the fuel system for a specified duration (for example, until a target fuel tank vacuum is reached) and then sealing the fuel system while monitoring any changes in fuel tank pressure (for example, a rate of vacuum drop or a final pressure reading). A fuel system leak can be identified based on the rate of vacuum release to atmospheric pressure, as described below. To perform the leak test, a vacuum generated at the intake manifold 44 can be applied to the fuel system. Specifically, the tank purge valve 112 and the tank vent valve 114 can be opened while the fuel tank vent valves 106A, 106B, and 108 remain open, so that a vacuum is drawn from the intake manifold 44 along the purge line 28. After reaching a threshold fuel tank vacuum, the tank purge valve and the tank vent valve can be closed while the tank vent valves remain open, and a fuel tank pressure release is monitored at the pressure sensor 120. Based on the pressure release rate (or vacuum drop rate) and the stabilized fuel tank outlet pressure after applying the engine intake vacuum, the presence of a fuel system leak can be determined.For example, in response to a vacuum discharge rate being faster than a threshold rate, a leak may be detected, and fuel system impairment may be indicated. However, if one of the fuel tank vent valves 106A, 106B, or 108 is temporarily blocked (i.e., accidentally closed) during the leak test, the fuel tank is isolated, and the fuel system volume is dramatically reduced. Since leak detection reference / threshold pressure values ​​are based on a fuel tank fill volume, the probability of a false leak detection increases if the fuel tank is isolated due to an unintentional temporary closure of a fuel tank vent valve. As explained below, under such conditions, a leak test may need to be aborted and repeated, so that only uncorrupted fuel system data can be used as the basis for leak identification. Referring again to Fig. 1, the vehicle system 6 can further include the control system 14. In the illustration, the control system 14 receives information from several sensors 16 (various examples of which are described here) and sends control signals to several actuators 81 (various examples of which are described here). For example, the sensors 16 can include the exhaust gas sensor 126, which is positioned upstream of the exhaust aftertreatment device, the temperature sensor 128, the MAP sensor 118, and the pressure sensor 129. Other sensors, such as additional pressure, temperature, air / fuel ratio, and composition sensors, can be connected to various points in the vehicle system 6. As another example, the actuators can include the fuel injector 66, the tank purge valve 112, the tank vent valve 114, and the throttle 62. The control system 14 can include a controller 12.The controller can receive input data from the various sensors, process the input data, and trigger the actuators in response to the processed input data based on instructions or programmed code according to one or more routines. Exemplary control routines are described here with reference to Figures 2 and 3. In this way, the system of Fig. 1 enables a method for a vehicle fuel system in which, during a fuel system leak test and in response to an accidental, temporary closure of a mechanical valve coupled to a fuel tank, the fuel system leak test is not terminated. Instead, the fuel system settings can be reset, and a fuel system leak test can be repeated. Now referring to Fig. 2, an exemplary routine 200 for applying a vacuum to a fuel system and identifying a fuel system leak based on a change in the fuel system pressure following the application of the vacuum is shown. If an unintentional, temporary closure (or blockage) of a fuel tank vent valve is identified during the fuel system leak test, the leak test is aborted and repeated at a later time to improve the reliability of the leak test results. At 202, it can be confirmed that the engine is running. For example, it can be confirmed that the vehicle is operating in an engine-on mode, with the vehicle being powered by the engine. If the engine is not running, at 203, engine-off leak test conditions can be confirmed. This can include confirming that a fuel tank temperature is within a threshold range, that a threshold time has elapsed since the engine was switched off, and that a threshold time has elapsed since the last leak test. Upon confirmation of engine-off leak test conditions, the routine at 204 includes performing an engine-off leak detection test. This involves identifying fuel system leaks by applying a natural vacuum to the fuel system with the engine off. Specifically, the fuel tank can be sealed when the engine is shut down by closing the tank purge valve and the tank vent valve. As the fuel tank cools, a vacuum is created in the vapor space of the fuel tank (due to the relationship between temperature and pressure of gases). The tank vent valve is then opened, and the rate of vacuum drop from the fuel tank is monitored. If the fuel tank pressure stabilizes at atmospheric pressure faster than expected, a fuel system leak is identified. When the engine is running, test 206 can determine whether the engine-on leak test conditions have been met. Leak detection approval conditions can include a wide variety of engine and / or fuel system operating conditions and parameters. Furthermore, leak detection approval conditions can include a wide variety of vehicle conditions. For example, engine-on leak detection eligibility conditions might include a fuel level in the fuel tank above a threshold, a temperature of one or more fuel system components within a predetermined temperature range (since temperatures that are too hot or too cold reduce leak detection accuracy), and a time / distance threshold elapsed since a previous leak test. In one example, a leak test might be performed after a vehicle has traveled a certain number of miles since a previous leak test or after a preset duration has passed since a previous leak test. If engine-on leak test eligibility conditions are not met, the routine might terminate. Upon confirmation of engine-in leak test conditions, a fuel system leak test can be initiated at 208. This involves opening a canister purge valve (CPV), allowing an engine intake manifold vacuum to be applied to the fuel system, specifically to the fuel tank via the canister. Simultaneously, a canister vent valve (CVV) can be closed to isolate the fuel system from the atmosphere. While the vacuum is applied, one or more passive tank vent valves (for example, valves 106A, 106B, and 108 in Fig. 1) can be considered open. The engine intake vacuum is then applied to the fuel system to increase the vacuum in the fuel tank, for example, to a threshold vacuum level (or for a threshold duration). This is thus also referred to as the vacuum increase phase of the fuel system leak test.As explained below, the fuel system can be isolated after the vacuum intensification phase, and a vacuum drop rate is monitored to identify leaks. Specifically, a fuel system leak is identified based on the rate at which the fuel tank vacuum is released to atmospheric pressure (also known as the vacuum release phase of the fuel system leak test). Thus, engine-on leak tests can be performed while a vehicle is in motion (for example, while a hybrid vehicle is moving in engine-on mode and the vehicle's constant speed is at steady-state speeds, for example, 40 mph). The inventors recognized that an unintentional, temporary closure of the fuel tank vent valves (the mechanical valves coupled to the fuel tank) can be caused by sources outside the fuel system, such as certain vehicle maneuvers performed while the vehicle is in motion.During vehicle maneuvers, such as sweeping left or right turns (e.g., cornering at speeds exceeding a threshold speed, and / or cornering at speeds exceeding threshold cornering speeds), uphill driving (e.g., driving along an incline steeper than a threshold gradient), and driving along a bumpy road (e.g., driving along a road surface with a level of slipperiness less than a threshold slipperiness), fuel may slosh over and briefly close the passive vent valves. Other maneuvers that can cause fuel to slosh over and a fuel tank vent valve to close include driving on undulating road surfaces, aggressive braking, and acceleration along any axis.If any of the fuel tank vent valves experience a momentary blockage while a leak test is being performed, the leak test results may be corrupted. Specifically, a momentary, unintentional closure of any of the passive fuel tank vent valves can cause the fuel tank to become isolated from the rest of the fuel system. This, in turn, reduces the volume of the fuel system. Since the thresholds used for the vacuum boost and / or drain phases of a leak test are a function of the fuel tank fill volume, if the fuel tank is isolated (due to the temporary vent valve blockage), a false-positive leak detection may occur, and incorrect diagnostic codes may be set. As a result, the leak diagnosis becomes less robust and less reliable.This can trigger an unjustified error message, possibly accompanied by a warning light, leading to unnecessary maintenance. To improve the reliability of leak test results, in response to an unintentional, temporary closure of a fuel tank vent valve during a given leak test, all data collected during the given leak test cycle can be disregarded, original (pre-test) fuel system settings can be resumed, and a leak test can be repeated (until a complete leak test can be performed without any indication of vent valve clogging). Referring again to Fig. 2, it can be determined at 210 during the vacuum intensification phase of the leak test whether vent valve blockages have been identified. In particular, it can be determined whether there has been any temporary, unintentional closure of one or more mechanical valves coupled to the fuel tank, unintentional closure due to sources outside the fuel system (as discussed above). It can also be determined whether the valves subsequently close after the brief, unintentional opening. Thus, the vent valves may be briefly and unintentionally opened and closed several times during the vacuum intensification phase of the leak test due to vehicle maneuvers performed during the leak test. As shown in Fig.3. If implemented, a controller can identify the unintentional, momentary opening and closing (and re-opening and re-closing) of the fuel tank vent valves during the vacuum intensification phase based on the presence of fuel tank pressure fluctuations during the vacuum intensification phase, the timing or location of the fluctuation points, and the vacuum change rate during the vacuum intensification phase. For example, the identification of unintentional, momentary closing of the mechanical valve during vacuum application to the fuel tank can be based on the presence of fluctuations in fuel tank pressure during vacuum application and / or on the fact that a vacuum intensification rate during vacuum application exceeds a (first) threshold rate.Similarly, the reopening of the temporarily closed vent valve during vacuum application can be indicated by fluctuations or a sudden decrease in the vacuum intensification rate. Thus, one or more (for example, multiple) fluctuations may occur during the vacuum application to the fuel tank, as described here. If the vent valve blockage is confirmed during vacuum intensification, then the routine at 218 includes interrupting and not terminating the leak test. Not terminating the fuel system leak test involves resuming the (original) fuel system settings from before the leak test was initiated. For example, the CVV may be opened, the CPV closed, and other fuel system valves that were set to the closed position during the leak test may be set to the open position (and vice versa). Returning the valves to their original (pre-test) settings allows the vacuum applied to the fuel tank to dissipate to atmospheric pressure conditions.Failure to complete the fuel system leak test also includes disregarding all pressure data collected during the application of vacuum to the fuel tank and the subsequent release of the vacuum, and not indicating a fuel system leak based on the vacuum release rate. Additionally, a diagnostic code may be set at 220 to indicate that the leak test was not completed due to a blocked fuel tank vent valve. Next, at 222, after resuming the original fuel system settings, the leak test can be repeated. Specifically, the CPV can be reopened, the CVV closed, and the fuel tank vent valves assumed to be open. Repeating the leak test still involves reapplying engine intake vacuum to the fuel tank while the mechanical fuel tank valves are assumed to be open, and after reapplying, remonitoring the vacuum build-up and subsequent vacuum release in the fuel tank. Thus, the routine can proceed to complete a fuel system leak test if vent valve clogging is identified during the subsequent vacuum build-up(s).If no vent valve blockage is present during the vacuum intensification phase of the leak test repetition, the routine can proceed to step 212 to re-isolate the fuel tank with the mechanical valve assumed to be open and monitor the vacuum release again. Then, based on the vacuum release during this remonitoring, the routine can indicate a fuel system leak if no further vent valve blockage is identified (as detailed below). Referring again to 210, if no vent valve blockage is identified during vacuum intensification (in an initial leak test attempt initiated at 208, or a subsequent leak test repetition initiated at 222), the routine proceeds to 212 to continue the vacuum release phase of the leak test. This routine involves, after applying a threshold vacuum level to the fuel tank, isolating the fuel tank by closing the CPV while keeping the CVV closed, and assuming the fuel tank vent valves are open. Subsequent release of the vacuum to atmospheric pressure is monitored. At 214, during the vacuum release phase of the leak test, it can be determined whether vent valve blockages have been identified. In particular, it can be determined whether there has been any temporary, unintentional closure of one or more mechanical valves coupled to the fuel tank, the unintentional closure being attributable to sources outside the fuel system (as discussed previously). Furthermore, it can be determined whether there has been a subsequent, unintentional opening of the valves back to their original settings during the vacuum release phase. Thus, the vent valves may be opened and closed briefly and unintentionally several times during the vacuum release phase of the leak test based on vehicle maneuvers performed during the execution of the leak test. As in connection with Fig.3. A controller can identify the short-term opening and closing (and re-opening and re-closing) of the fuel tank vent valves during the vacuum discharge phase based on the presence of fuel tank pressure fluctuations during the vacuum discharge phase, the timing or location of the fluctuation points, and the vacuum change rate during the vacuum discharge phase. For example, the identification of an unintentional, temporary closure of the mechanical valve during vacuum discharge from the isolated fuel tank can be based on the presence of fuel tank pressure fluctuations during the isolation of the fuel tank and / or on the fact that a vacuum discharge rate during the isolation of the fuel tank exceeds a (second, different) threshold rate.Similarly, the reopening of the temporarily closed vent valve during fuel tank isolation may be indicated by pressure fluctuations or a sudden decrease in the vacuum release rate. Thus, one or more (for example, multiple) fluctuations may be observed during the vacuum release phase, as described here. If a blocked vent valve is confirmed during vacuum release, the routine returns to 218 to interrupt the leak test. Additionally, a diagnostic code can be set to indicate that the leak test was not completed due to a blocked fuel tank vent valve. As previously described, the original (pre-test) fuel system settings can be restored, vacuum release to atmospheric pressure can be enabled, and pressure data collected up to that point during the leak test can be disregarded, thus preventing any fuel system leak from being identified due to the vacuum release. The fuel leak test can then be repeated (at 222). Therefore, the routine can continue to interrupt the completion of a fuel system leak test if a blocked vent valve is identified during vacuum release. If no vent valve blockage occurs during vacuum release during the repetition of the leak test, the routine can proceed to step 216 to terminate the leak test. During this process, the vacuum release can be monitored while the fuel tank is insulated, and a fuel system leak can be identified based on the vacuum release rate (for example, based on the fact that the vacuum release rate exceeds a threshold rate). In some embodiments, the leak opening size can also be determined based on a deviation of the monitored vacuum release rate from the threshold rate. In this way, leak detection based on vacuum release in an insulated fuel tank can only be identified if no unintentional, temporary closure of a fuel tank vent valve is detected during either the fuel tank vacuum intensification or vacuum release phase of the leak test. By interrupting a leak test if momentary tank vent valve clogging is identified and repeating the leak test, false leak detections can be reduced and leak test reliability can be improved. Referring to Fig. 3, an exemplary routine 300 for identifying a fuel tank vent valve blockage is shown. This routine identifies an unintentional and temporary closure of a mechanical vent valve coupled to the fuel tank during either a vacuum intensification or discharge phase of a leak test based on changes in the vacuum intensification or discharge rates and the presence of pressure fluctuations during the vacuum intensification or discharge phase. As explained below, opening of the temporarily closed vent valve and an unintentional and temporary re-closing of the vent valve during the vacuum intensification or discharge phase can also be determined. Exemplary changes in the vacuum discharge or discharge rates that can be used to derive vent valve blockages are subsequently shown in Figs. 5-11 and illustrated with a diagram in Fig.The 4 expected pressure profiles shown were compared. At 302, the routine includes confirming a vacuum intensification phase of the fuel system leak test. For example, it can be confirmed that the canister purge valve is open, that a canister purge valve is closed, and that an engine intake vacuum is applied to the fuel system (specifically, via the canister to the fuel tank). Furthermore, a vacuum intensification rate can be monitored during the application of the engine intake vacuum. For example, based on engine operating conditions, the magnitude of the engine intake vacuum generated and applied to the fuel system via a purge line (also referred to as the purge rate) can be estimated. At 304, it can be determined whether the vacuum intensification rate is faster than a threshold rate. In one example, the threshold rate could be based on the estimated purge rate. If the vacuum intensification rate is higher than the threshold rate, then at 308 it can be determined that the vacuum in the fuel tank is increasing faster than expected due to an unintentional and temporary closure of a mechanical fuel tank vent valve. In other words, it can be determined that a fuel tank vent valve is momentarily blocked due to sources outside the fuel system, such as sudden and drastic vehicle maneuvers. At 310, it can be determined whether there is a fluctuation in the vacuum gain rate. For example, it can be determined whether there is a sudden change (for example, a sudden decrease) in the vacuum gain rate. If so, then at 312, it can be determined that the fuel tank vent valve has been released. That is, the accidentally closed vent valve has reopened. Otherwise, it can be determined that the valve is still blocked. In one example, during the vacuum gain phase of the leak test, an initial vacuum gain rate might be as expected. Then, in the middle of the vacuum gain phase, the vacuum gain rate might suddenly increase, indicating a temporary closure of the vent valve.After a period of vacuum amplification at the increased rate, the vacuum may fluctuate, and the vacuum amplification rate may drop back to the initial rate, indicating a reversal of the temporary valve closure. In another example, the vacuum amplification rate may be elevated at the beginning of the vacuum amplification phase. Thus, any one or more fuel tank vent valves may be inadvertently closed and reopened multiple times during the vacuum intensification phase. This means that multiple fluctuations may occur during the vacuum intensification phase. Therefore, following the determination of the vent valve release at 312 to 304, the routine can return to determine whether the vent valve blockage occurred again during the vacuum intensification phase. If the vacuum intensification at 304 is not faster than the threshold rate, it can be determined at 306 that no unintentional, temporary closure of the fuel tank vent valves has occurred during the vacuum intensification phase of the leak test, and the leak test can proceed to the vacuum release phase. The next step in the 320 routine involves confirming a vacuum release phase of the fuel system leak test. For example, it can be confirmed that the tank purge valve is closed, the tank vent valve is closed, the fuel tank is isolated, and that a threshold level of engine intake vacuum has already been applied to the fuel system. It can also be confirmed that the fuel tank vacuum is at (or above) a threshold level. Furthermore, a vacuum release rate during fuel tank isolation can be monitored. For example, based on engine operating conditions, fuel system conditions, and ambient temperature and pressure conditions, an expected rate of fuel tank vacuum release to atmospheric pressure can be estimated. At 322, it can be determined whether the vacuum gain rate is faster than a threshold rate. In one example, it can be determined whether the vacuum gain rate is faster than both a first threshold rate (where the first threshold rate is based on the expected vacuum discharge rate from the fuel tank in the absence of any fuel system leaks) and a second threshold rate (where the second threshold rate is based on an expected vacuum discharge rate in the presence of a fuel system leak and is potentially different from the first threshold rate). In one example, the second threshold rate may be higher than the first threshold rate. If the vacuum discharge rate is higher than both the first and second threshold rates, then at 326 it can be determined that the vacuum is being discharged from the fuel tank faster than expected due to an unintentional and temporary closure of a mechanical fuel tank vent valve.In other words, it can be determined that a fuel tank vent valve is temporarily blocked due to sources outside the fuel system, such as sudden and drastic vehicle maneuvers. This is because the vacuum release rate during a fuel tank vent valve blockage event is significantly greater than a vacuum release rate due to a leak in the fuel system. At 328, it can be determined whether there is a fluctuation in the vacuum release rate. For example, it can be determined whether there is a sudden change (for example, a sudden decrease) in the vacuum release rate. If so, then at 330, it can be determined that the fuel tank vent valve has been released. That is, the vent valve that was accidentally closed has reopened. In one example, during the vacuum release phase of the leak test, an initial vacuum release rate may be as expected. Then, in the middle of the vacuum release phase, the vacuum release rate may suddenly increase, indicating a temporary closure of the vent valve. After a period of vacuum release at the increased rate, the vacuum may fluctuate, and the vacuum release rate may decrease back to the initial release rate, indicating a reversal of the temporary valve closure.In alternative examples, the vacuum release rate can increase at the beginning of the vacuum release phase, such as during a transition from the vacuum intensification phase to the vacuum release phase. Thus, one or more fuel tank vent valves may be inadvertently closed and then reopened multiple times during the vacuum release phase. This means that multiple fluctuations can occur during the vacuum release phase. Therefore, after determining the vent valve release at 330 to 322, the routine can return to determine if the vent valve blockage has recurred during the vacuum release phase. If, at 322, the vacuum release is not faster than the threshold rate, then at 324 it can be determined that no accidental, temporary closure of the fuel tank vent valves occurred during the vacuum release phase of the leak test. It is understood that while the above routine demonstrates the identification of a fuel tank vent valve blockage based on a vacuum intensification or discharge rate during a leak test and the identification of a fuel tank vent valve release based on a fuel tank pressure fluctuation, in other embodiments the blockage and release of the vent valve can be directly derived from fluctuations in fuel tank pressure occurring during a vacuum intensification and / or vacuum discharge phase of a leak test.Furthermore, it is understood that while the above routine demonstrates the identification of an unintentional, temporary closure of the mechanical valve based on one or more fluctuations in fuel tank pressure during the application of vacuum to the fuel tank or one or more fluctuations in fuel tank pressure during the release of the vacuum, in further embodiments the identification of an unintentional, temporary closure of the mechanical valve may be based on one or more fluctuations in fuel tank pressure during a fuel tank pressure stabilization period. Referring to Figures 4-11, exemplary changes in fuel tank vacuum level during the vacuum intensification and release phases of a leak test are shown. In particular, Figure 4 shows an example of a leak test in which no fuel tank vent valve blockage occurs during the leak test, whereas in the examples shown in Figures 5-11, one or more fuel tank vent valve blockages occur during the vacuum intensification and / or release phases. The characteristic map 400 of Fig. 4 shows an exemplary change in the fuel tank vacuum at plot 402 during a vacuum intensification and release phase of a leak test. During the vacuum intensification phase, vacuum (from an engine inlet) is applied to the fuel tank to increase the vacuum in the fuel tank to a target or threshold vacuum level 401. For example, a tank purge valve can be opened to allow an engine inlet vacuum to be applied to the fuel tank and the fuel tank pressure to drop to the threshold vacuum level. Then, during the vacuum release phase, the fuel tank can be isolated, and a rate of release of the vacuum to atmospheric pressure is monitored. For example, the tank purge valve can be closed to allow the fuel tank vacuum to drop from the threshold vacuum level.If there is no leak in the fuel system, fuel tank vacuum can be discharged at a threshold rate, as shown by plot 402 (solid line). However, if there is a leak in the fuel system, fuel tank vacuum can be discharged at a rate faster than the threshold rate, as shown by plot 403 (dashed line). Characteristic map 500 of Fig. 5 shows another exemplary change in the fuel tank vacuum at plot 502 during a vacuum intensification and release phase of a leak test. When vacuum is applied to the fuel tank to increase the vacuum to a swelling vacuum level 501, the vacuum intensification rate is higher than the swelling rate for a certain period during the vacuum intensification phase (as can be seen by comparing the slope of plot 402 with the slope of plot 502 during the vacuum intensification phase of each leak test). In particular, segment 503 shows a region where, at the beginning of the vacuum intensification phase, the vacuum is intensified at an increased rate. As a result of the rate being higher than the swelling vacuum intensification rate, it can be determined that a fuel tank vent valve has been inadvertently and temporarily closed (i.e., a passive tank vent valve is blocked).Furthermore, the high vacuum gain rate can cause a line-blocked code and vacuum overshoot. For example, the controller may include additional logic to confirm that a line is blocked. Then, in response to a sudden fluctuation in fuel tank vacuum, it may be determined that the vent valve has been released. In particular, segment 504 shows a region in the middle of the vacuum gain phase where the vacuum abruptly changes from a state higher than the threshold rate to a state lower than the threshold rate and gradually returns to an expected pressure profile. In the example shown, the valve remains released into and during the vacuum release phase. As in Fig.If step 2 is performed, the leak test may be interrupted in response to an indication of a vent valve blockage, and the fuel tank vacuum drain rate from plot 502 cannot be used to identify a fuel system leak. Instead, a leak test may be repeated. Characteristic map 600 of Fig. 6 shows another exemplary change in the fuel tank vacuum at plot 602 during a vacuum intensification and release phase of a leak test. While the engine intake vacuum is applied to the fuel tank to increase the vacuum to the threshold vacuum level 601, here, in the middle of the vacuum intensification phase, the vacuum intensification rate is higher than a threshold rate for a certain duration (as can be seen by comparing the slope of plot 602 (solid line) with the slope of plot 402 (Fig. 4) during the vacuum intensification phase of each leak test). In particular, segment 606a shows an initial region of the vacuum intensification phase in which the vacuum is increased faster than a threshold rate, essentially at a purge line vacuum level 607, while an actual fuel tank vacuum (plot 604, dotted line) is much lower.Here, the vacuum gain rate is proportional to the volume evaluated for a given manifold vacuum. In response to the rate being higher than the throttling vacuum gain rate, it can be determined that at 606a, a first unintended fuel tank vent valve closure occurred (that is, the valve was blocked for the first time). Then, in response to a sudden fluctuation in the fuel tank vacuum, it can be determined that the vent valve was released. In particular, segment 608a shows an initial region of the vacuum gain phase in which the vacuum suddenly changes from a value higher than the throttling rate to a value lower than the throttling rate and gradually returns to an actual fuel tank vacuum level. Segment 606b shows a second area in the middle of the vacuum amplification phase where the vacuum is amplified faster than the threshold rate. In response to the rate being higher than the threshold vacuum amplification rate, it can be determined that a second unintended closure of the fuel tank vent valve has occurred (i.e., the valve has become clogged a second time). Here, the valve may remain clogged for the remainder of the vacuum amplification phase. Thus, depending on the timing of the clogging event, the high vacuum amplification rate can also trigger a line-blocked code or a large vacuum release. In one example, the controller may include additional logic to confirm the existence of a blocked line. During a transition to the vacuum release phase of the leak test in response to a sudden fluctuation in the fuel tank vacuum, it can then be determined that the vent valve has been released.In particular, segment 608b shows an initial region at the beginning of the vacuum discharge phase where the rate of change of the vacuum suddenly fluctuates and approaches fuel tank vacuum. Thus, in the example shown, the valve clogs several times during the vacuum intensification phase and is released at the beginning of the vacuum discharge phase. As illustrated in Fig. 2, the leak test may be interrupted in response to the indication of (repeated) vent valve clogging and release; the leak test may be aborted, and the fuel tank vacuum discharge rate from plot 602 cannot be used to identify a fuel system leak. Instead, a leak test can be repeated. Characteristic map 700 of Fig. 7 shows another exemplary change in the fuel tank vacuum at plot 702 during a vacuum boost and release phase of a leak test. Here, multiple fuel tank vent valve plugging and unblocking occur during the vacuum boost phase, and the vent valve remains unblocked during the vacuum release phase. In particular, when engine intake vacuum is applied to a fuel tank to boost the vacuum to a swell vacuum level 701, the vacuum boost rate is higher than the swell rate for a duration in the middle of the vacuum boost phase. Segment 706a shows an initial region of the vacuum boost phase in which the vacuum is boosted more rapidly than the swell rate, essentially at a purge line vacuum level 707, while an actual fuel tank vacuum (plot 704, dotted line) is much lower.In response to the rate being higher than the swelling vacuum gain rate, it can be determined that an initial unintended fuel tank vent valve closure occurred at 706a. Then, in response to a sudden fluctuation in fuel tank vacuum, it can be determined that the vent valve was released. Specifically, segment 708a shows an initial region of the vacuum gain phase, in which the vacuum changes from a state higher than the swelling rate to a state lower than the swelling rate, and gradually returns to a fuel tank vacuum level. Segment 706b shows a second region of the vacuum amplification phase, in which the vacuum is amplified more rapidly than the swell rate. In response to the rate being higher than the swell vacuum amplification rate, it can be determined that a second unintended fuel tank vent valve closure has occurred (that is, the valve has been clogged a second time). Then, in response to a sudden fluctuation in the fuel tank vacuum, it can be determined that the vent valve has been released. Specifically, segment 708b shows a second region of the vacuum amplification phase, in which the rate of change of the vacuum suddenly fluctuates and approaches the fuel tank vacuum. Thus, in the example shown, the valve clogs several times during the vacuum amplification phase and remains released during the vacuum discharge phase. This behavior allows for normal vacuum discharge. As shown in Fig.2. However, if the leak test is performed in response to an indication of (repeated) clogging and unblocking of the vent valve, the leak test may be interrupted, and the fuel tank vacuum drain rate from plot 702 (even if normal) cannot be used to identify a fuel system leak. Instead, a leak test can be repeated. Characteristic map 800 of Fig. 8 shows another exemplary change in the fuel tank vacuum at plot 802 during a vacuum intensification and release phase of a leak test. When vacuum is applied to the fuel tank to increase the vacuum to the thriving vacuum level 801, the vacuum intensification rate is higher than the thriving rate for a duration (as can be seen by comparing the slope of plot 402 with the slope of plot 802 during the vacuum intensification phase of each leak test) during the vacuum intensification phase. In particular, segment 803 shows a region in the middle of the vacuum intensification phase where the vacuum is increased more rapidly than the thriving rate. As a result of the rate being higher than the thriving vacuum intensification rate, it can be determined that a fuel tank vent valve has been inadvertently and temporarily closed (i.e., the valve has become blocked).Furthermore, the high vacuum gain rate can trigger a line-blocked code and cause vacuum overshoot. In one example, the controller might include additional logic to confirm that a line is blocked. The valve then remains blocked during the vacuum gain phase. During a transition to the vacuum release phase, a sudden fluctuation in the fuel tank vacuum occurs, indicating that the vent valve has been released. In particular, segment 804 shows a region at the beginning of the vacuum gain phase where the vacuum abruptly changes from a state higher than the threshold rate to a state lower than the threshold rate. As shown in Fig.If step 2 is performed, the leak test may be interrupted in response to the indication of a blocked and released vent valve, and the fuel tank vacuum drain rate from plot 802 cannot be used to identify a fuel system leak. Instead, a leak test may be repeated. Characteristic map 900 of Fig. 9 shows another exemplary change in the fuel tank vacuum at plot 902 during a vacuum intensification and release phase of a leak test. When vacuum is applied to the fuel tank to increase the vacuum to a swelling vacuum level 901, the vacuum intensification rate is higher than the swelling rate for a duration (as can be seen by comparing the slope of plot 402 with the slope of plot 902 during the vacuum intensification phase of each leak test) during the vacuum intensification phase. In particular, segment 903 shows a region in the middle of the vacuum intensification phase where the vacuum is increased faster than the swelling rate. As a result of the rate being higher than the swelling vacuum intensification rate, it can be determined that a fuel tank vent valve has been inadvertently and temporarily closed (i.e., the valve has become blocked).Here, the valve remains blocked for the remainder of the vacuum intensification phase as well as during the subsequent vacuum discharge phase. This behavior could therefore lead to a false pass if there is a leak on the tank side and / or trigger a line-blocked code. As illustrated in Fig. 2, the leak test may be interrupted in response to the indication of a blocked vent valve, and the fuel tank vacuum discharge rate from plot 902 cannot be used to identify a fuel system leak. Instead, a leak test can be repeated. Characteristic map 1000 of Fig. 10 shows another exemplary change in the fuel tank vacuum at plot 1002 during a vacuum intensification and release phase of a leak test. When vacuum is applied to the fuel tank to intensify the vacuum to a throttling vacuum level 1001, the vacuum intensification rate is higher than the throttling rate for a duration (as can be seen by comparing the slope of plot 402 with the slope of plot 1002 during the vacuum intensification phase of each leak test) during the vacuum intensification phase. In particular, segment 1003 shows a region in the middle of the vacuum intensification phase where the vacuum is intensified faster than the throttling rate. As a result of the rate being higher than the throttling vacuum intensification rate, it can be determined that a fuel tank vent valve has been inadvertently and temporarily closed (i.e., the valve has become blocked).Furthermore, the vacuum gain rate can trigger a line-blocked code and cause a vacuum overshoot. The valve then remains blocked for the remainder of the vacuum gain phase and during the transition to the vacuum release phase. Midway through the vacuum release phase, a sudden fluctuation in the fuel tank vacuum occurs, indicating that the vent valve has been released. Specifically, segment 1004 shows a region midway through the vacuum release phase where the vacuum fluctuates suddenly. As illustrated in Fig. 2, the leak test may be interrupted in response to the indication of the vent valve blockage and release, and the fuel tank vacuum release rate from plot 1002 cannot be used to identify a fuel system leak. Instead, a leak test can be repeated. Characteristic map 1100 of Fig. 11 shows another exemplary change in the fuel tank vacuum at plot 1102 during a vacuum intensification and release phase of a leak test. When vacuum is applied to the fuel tank to increase the vacuum to a throttling vacuum level 1101, the vacuum intensification rate is higher than the throttling rate for a certain duration during the vacuum intensification phase (as can be seen by comparing the slope of plot 402 with the slope of plot 802 during the vacuum intensification phase of each leak test). In particular, segment 1103 shows that from the start of the vacuum intensification, the vacuum is increased more rapidly than the throttling rate. As a result of the rate being higher than the throttling vacuum intensification rate, it can be determined that a fuel tank vent valve has been inadvertently and temporarily closed (i.e., the valve has become blocked).Furthermore, the high vacuum gain rate can trigger a line-blocked code and cause a vacuum overshoot. The valve may then remain blocked during the vacuum gain phase. During the transition to the vacuum release phase, a sudden fluctuation in the fuel tank vacuum occurs, indicating that the vent valve has been released. In particular, segment 1104 shows a region at the beginning of the vacuum gain phase where the vacuum fluctuates suddenly. As illustrated in Fig. 2, the leak test may be interrupted in response to the indication of the vent valve blockage and release, and the fuel tank vacuum release rate from plot 1102 cannot be used to identify a fuel system leak. Instead, a leak test can be repeated. Referring to Fig. 12, exemplary fuel system leak test operations are shown. Map 1200 shows the status of a fuel system leak test (on or off) at plot 1202, the status of a tank purge valve (open or closed) coupled between an engine intake manifold and a fuel system tank at plot 1204, the status of a fuel tank vent valve (open or closed / blocked) at plot 1206, and changes in fuel pressure at plot 1208. Prior to t1, a vehicle engine may be running to power the vehicle. The tank purge valve (CPV) may be closed because engine-on leak test conditions are not met. At t1, in response to the fulfillment of leak test conditions, an initial leak test may be initiated (Figure 1202), and the tank purge valve may be opened (Figure 1204) to increase the engine inlet vacuum in the fuel tank. Thus, one or more passive fuel tank valves may be expected to be open during the leak test (Figure 1206). As a vacuum is applied to the fuel tank, the fuel tank pressure (Figure 1208) may begin to decrease. A vacuum increase phase of the leak test may be performed between t1 and t2, reducing the fuel tank pressure to a target pressure level (or target vacuum level). During this first vacuum intensification (between t1 and t2) a fuel tank pressure fluctuation may be experienced, as shown in 1210. In particular, the vacuum gain rate can be increased midway through the vacuum gain phase (that is, vacuum is drawn at a higher rate than a threshold rate). The increased vacuum gain rate can be maintained for a period of time during the vacuum gain. Then, the vacuum gain rate can just as suddenly revert to the original vacuum gain rate (which is lower than the threshold rate). In response to the pressure fluctuation during this initial vacuum gain, an unintentional, temporary closure and subsequent reopening of the tank vent valve may be indicated (Figure 1206). To reduce the possibility of a false leak detection, the tank purge valve can be closed at t2 in response to the indication (Figure 1204), and a tank vent valve can be opened to release vacuum from the fuel tank.Consequently, fuel tank pressure (between t2 and t3) can be released and stabilize at atmospheric pressure (Figure 1208). Furthermore, the leak test can be interrupted (Figure 1202). In particular, an engine control unit cannot identify fuel system leaks based on the first fuel tank vacuum release (between t2 and t3) immediately after the first vacuum intensification (between t1 and t2). After completion of the first fuel tank vacuum drain, a leak test can be restarted at t3 (Figure 1202). At t3, the purge valve can be reopened (Figure 1204) to reinforce the engine intake vacuum in the fuel tank (Figure 1208). The vacuum intensification phase of the leak test can continue from t3 to t4 to draw a threshold vacuum level at the fuel tank. No fuel tank pressure fluctuations can be observed during the vacuum intensification between t3 and t4. Accordingly, it can be determined that the fuel tank vent valves (which were assumed to be open) are open and no valve blockage has occurred (Figure 1206). In response to the absence of pressure fluctuations during the vacuum intensification of the restarted leak test, the purge valve can be closed at t4 to isolate the fuel tank and initiate a vacuum drain phase of the leak test.Accordingly, fuel tank pressure can be released to atmospheric pressure. A fuel system leak can then be identified based on the vacuum release rate between t4 and t5. Specifically, if the vacuum release rate is slower than a threshold rate (plot 1208), it can be determined that there is no fuel system leak. Conversely, if the vacuum release rate is faster than a threshold rate (plot 1212, dashed line), it can be determined that there is a fuel system leak. In this way, a fuel system leak test can be terminated, and a fuel system leak can only be identified based on a vacuum drain rate if no pressure fluctuations occur during the leak test. By disregarding fuel tank vacuum drain data in the event of a fuel tank vent valve blockage, and furthermore, by resuming original fuel system settings for a subsequent leak test attempt, false-positive leak detections caused by pressure fluctuations can be reduced. It is understood that while the above example demonstrates the identification of a fuel tank vent valve blockage based on pressure fluctuations experienced during a vacuum build-up phase, in other examples a fuel tank vent valve blockage can also be identified based on pressure fluctuations experienced during a vacuum release phase. For instance, a controller may open a purge valve to build engine intake vacuum into a fuel tank and, after the vacuum is established, close the purge valve to isolate the fuel tank and vent the vacuum. In response to a pressure fluctuation during an initial vacuum release, the controller may not be able to identify fuel system leaks based on the initial vacuum release. Furthermore, the controller may indicate an unintentional, temporary closure of a mechanical vent valve coupled to the fuel tank.In contrast, if no pressure fluctuation occurs during a second vacuum release, the control unit can identify fuel system leaks based on the second vacuum release. Furthermore, if a pressure fluctuation occurs during the first vacuum release, and after the vacuum has been released, the control unit can reopen the purge valve to increase the engine intake vacuum in the fuel tank. After the vacuum increase, it can close the purge valve to re-insulate the fuel tank. If no pressure fluctuation occurs during the vacuum increase, the control unit can identify fuel system leaks based on the third vacuum release immediately after the vacuum increase. In this way, pressure fluctuations and changes in vacuum gain and discharge rates observed during a leak test can be correlated with a temporary and accidental closure of a fuel tank vent valve due to external sources, such as vehicle maneuvers. By disregarding leak test data when a valve blockage occurs and repeating the leak test, increased vacuum discharge rates resulting from the temporary valve blockage cannot be incorrectly identified as a fuel system leak. Repeating the leak test improves the reliability and accuracy of fuel system leak diagnosis. It should be noted that the example control routines contained herein can be used with various engine and / or vehicle system configurations. The specific routines described here can represent one or more of a number of processing strategies, such as event-driven, interrupt-driven, multitasking, multithreading, and the like. Thus, various actions, operations, or functions shown can be performed in the sequence shown, in parallel, or, in some cases, omitted. Furthermore, the described actions can graphically represent code to be programmed into the computer-readable storage medium in the engine control system.

Claims

A method for a fuel system (18) for a vehicle, comprising: during a fuel system leak test and in response to an unintentional, temporary closing of a mechanical valve (106A, 106B, 108) coupled to a fuel tank (20), not terminating the fuel system leak test, characterized in that the fuel system leak test comprises: closing a tank vent valve (114) coupled between a fuel vapor reservoir (22) and the atmosphere; opening a purge valve (112) coupled between a fuel vapor reservoir (22) and an engine inlet; applying an engine inlet vacuum to the fuel tank (20) with the mechanical valve (106A, 106B, 108) open;and after installation, isolating the fuel tank (20) by closing the purge valve (112) while the mechanical valve (106A, 106B, 108) is still open, monitoring a vacuum drain in the fuel tank (20) and indicating a fuel system leak based on a vacuum drain rate.; Method according to claim 1, wherein an identification of an unintentional, temporary closing of the mechanical valve (106A, 106B, 108) is based on one or more fluctuations in the fuel tank pressure during the application of the vacuum to the fuel tank (20) or one or more fluctuations in the fuel tank pressure during the vacuum discharge. Method according to claim 2, wherein the identification is further based on a vacuum intensification rate during the application of the vacuum to the fuel tank (20) and a vacuum release rate during the insulation of the fuel tank (20). The method of claim 3, wherein the identification comprises indicating an unintentional, temporary closing of the mechanical valve (106A, 106B, 108) during the application of vacuum to the fuel tank (20) based on the fact that the rate of vacuum intensification in the fuel tank (20) is higher than a first threshold rate, and indicating an unintentional, temporary closing of the mechanical valve (106A, 106B, 108) during the isolation of the fuel tank (20) based on the fact that the rate of vacuum release in the fuel tank (20) is higher than a second, different threshold rate. Method according to claim 4, wherein the identification further comprises indicating the reopening of the temporarily closed mechanical valve (106A, 106B, 108) during the application of the vacuum based on a change in the vacuum gain rate and indicating the reopening of the temporarily closed mechanical valve (106A, 106B, 108) during the isolation of the fuel tank (20) based on a change in the vacuum discharge rate. The method of claim 1, wherein the non-termination of the fuel system leak test comprises closing the purge valve (112), opening the tank vent valve (114) and resuming the fuel system settings from before the initiation of the leak test and the non-indication of a fuel system leak based on the vacuum drain rate during monitoring. The method according to claim 6, further comprising, after resuming the fuel system settings, reopening the purge valve (112), closing the tank vent valve (114), reapplying the engine inlet vacuum to the fuel tank (20) with the mechanical valve (106A, 106B, 108) open, after reapplying the vacuum, re-insulating the fuel tank (20) by closing the purge valve (112) with the mechanical valve (106A, 106B, 108) still open, re-monitoring the vacuum release, and, in response to no unintentional, temporary closing of the mechanical valve (106A, 106B, 108), indicating a fuel system leak based on the vacuum release rate during the re-monitoring. Method according to claim 1, wherein the fuel system leak test is performed while the vehicle is in motion, and the unintentional, temporary closing of the mechanical valve (106A, 106B, 108) is caused by sources located outside the fuel system (18), including vehicle maneuvers performed while the vehicle is in motion. The method of claim 8, wherein the vehicle maneuvers comprise vehicle cornering at vehicle speeds exceeding a threshold speed, vehicle cornering at speeds exceeding a threshold cornering speed, vehicle driving along a gradient higher than a threshold gradient, and vehicle driving along a road surface with a smoothness lower than a threshold smoothness. A method for a vehicle fuel system, characterized in that the method comprises: opening a purge valve (112) to use the engine intake vacuum to increase the vacuum in the fuel tank (20); in response to a pressure fluctuation during a first vacuum increase, closing the purge valve (112), venting the vacuum from the fuel tank (20), and failing to identify fuel system leaks based on a first fuel tank vacuum release immediately after the first vacuum increase; and in response to no pressure fluctuation during a second vacuum increase, closing the purge valve (112) to isolate the fuel tank (20) and identifying fuel system leaks based on a second fuel tank vacuum release immediately after the second vacuum increase. The method of claim 10, further comprising opening the purge valve (112) to apply engine inlet vacuum to the fuel tank (20) during a third vacuum intensification in response to pressure fluctuations during the first vacuum intensification and after the first fuel tank vacuum drain, and closing the purge valve (112) to isolate the fuel tank (20) after the third vacuum intensification, and identifying fuel system leaks based on vacuum drain immediately after the variable vacuum intensification number in response to no pressure fluctuations during a discrete number of vacuum intensifications. The method according to claim 11, which further comprises, in response to pressure fluctuations during the first vacuum intensification, indications of an unintentional, temporary closing of a mechanical container venting valve (114) coupled to the fuel tank (20). A method for a fuel system coupled to a vehicle engine (10), characterized in that the method comprises: opening a purge valve (112) to increase the engine inlet vacuum in a fuel tank (20); after applying the vacuum, closing the purge valve (112) to isolate the fuel tank (20) and vent the vacuum; in response to a pressure fluctuation during a first vacuum release, not identifying any fuel system leaks based on the first vacuum release; and in response to no pressure fluctuations during a second vacuum release, identifying any fuel system leaks based on the second vacuum release. The method according to claim 13, further comprising, in response to pressure fluctuations during the first vacuum release and after venting the vacuum, reopening the purge valve (112) to reinforce the engine inlet vacuum in the fuel tank (20), closing the purge valve (112) after the vacuum reinforcement to reinsulate the fuel tank (20), and, in response to no pressure fluctuation during the vacuum reinforcement, identifying fuel system leaks based on a third vacuum release immediately after the vacuum reinforcement. The method of claim 14, which further comprises indicating, in response to pressure fluctuations during the first vacuum release, an unintentional, temporary closing of a mechanical venting valve (114) coupled to the fuel tank (20). Fuel system for a vehicle comprising: an engine (10) containing an intake manifold (44); a fuel tank (20) coupled to the intake manifold (44) via a fuel vapor reservoir (22), the fuel tank (20) containing a mechanical tank venting valve (114); a purge valve (112) coupled between the intake manifold (44) and the fuel vapor reservoir (22), configured to allow the application of an intake manifold vacuum to the fuel tank (20) via the fuel vapor reservoir (22); a tank venting valve (114) coupled to the fuel vapor reservoir (22), configured to isolate the fuel system (18) from the atmosphere; and a control (12) with computer-readable instructions for: closing the tank venting valve (114);characterized in that the control unit (12) is also configured to: open the purge valve (112) when the mechanical tank vent valve (114) is assumed to be open, in order to increase a threshold level of the inlet manifold vacuum at the fuel tank (20); after the vacuum increase, close the purge valve (112) to release vacuum from the fuel tank (20); in response to a pressure fluctuation during the vacuum increase, indicate an unintentional, temporary closure of the mechanical tank vent valve (114) and fail to identify a fuel system leak based on the vacuum release; and in response to no pressure fluctuation during the vacuum release, identify a fuel system leak based on the vacuum release. Fuel system (18) according to claim 16, wherein the failure to identify a fuel system leak based on the vacuum drain comprises disregarding data from the vacuum drain, closing the purge valve (112) and resuming the fuel system settings from before the vacuum boost. Fuel system (18) according to claim 17, which further comprises, after resumption of the fuel system settings, reopening the purge valve (112) to reapply the threshold height of the inlet manifold vacuum to the fuel tank (20), closing the tank vent valve (114) again, and, in response to no pressure fluctuations during the reapplying of the vacuum, closing the purge valve (112) to release the vacuum from the fuel tank (20) and identifying a fuel system leak based on the vacuum release after the reapplying of the vacuum. Fuel system (18) according to claim 18, wherein the indication of an unintentional, temporary closing of the mechanical tank vent valve (114) comprises indicating a temporary closing of the tank vent valve (114) due to a vehicle maneuver.

Images