Self-learning algorithm for liquid level sensor diagnosis

Abstract

A fluid storage system includes a fluid tank with a level sensor and an injector for delivering fluid from the fluid tank in a controlled amount. A method for providing level sensor diagnostics for a variety of tank shapes, tank sizes, and fluid sensor types includes learning sensor resolution, fluid tank geometry, and fluid tank volume based on an output from a level sensor and an estimate of fluid delivered from a fluid tank.

Description

[0001] Cross-references to related applications This application claims the benefit and priority of U.S. Provisional Patent Application No. 63 / 535,807, filed August 31, 2023, which is incorporated herein by reference. Technical Field

[0002] This invention relates generally to level sensors, and more specifically, but not exclusively, to self-learning algorithms for diagnosing level sensors used in fluid storage systems for internal combustion engines and vehicles. Background Technology

[0003] Fluid storage systems used in engines and vehicles can be configured with a wide variety of level sensor types, tank sizes, and tank geometries. Diagnostics required for engines and vehicles may involve determining whether a level sensor is malfunctioning or otherwise providing incorrect readings. Existing techniques for level sensor diagnostics may limit the types of level sensors that can be used, restrict tank sizes, and / or limit tank geometries. Existing level sensor diagnostics may also generate erroneous sensor fault determinations, allowing only certain types of sensor faults to be detected, and / or may involve unique software and calibration builds for each unique tank configuration. While offering some benefits, existing methods face several challenges, drawbacks, limitations, and unresolved issues. Therefore, there remains a significant need for the devices, methods, and systems disclosed herein.

[0004] Disclosure of Example Implementations To clearly, concisely, and accurately describe exemplary embodiments of the invention, the ways and processes of making and using these exemplary embodiments, and to enable the practice, making, and use of these exemplary embodiments, reference will now be made to certain exemplary embodiments, including those illustrated in the accompanying drawings, and these exemplary embodiments will be described using specific language. However, it should be understood that this does not constitute a limitation on the scope of the invention, and that the invention includes and protects such changes, modifications, and further applications of the exemplary embodiments as would be apparent to those skilled in the art. Summary of the Invention

[0005] A fluid storage system is disclosed, comprising a fluid tank with a level sensor and an injector for delivering fluid from the fluid tank in a controlled manner. Level sensor diagnostics are provided for various tank shapes, tank sizes, and fluid sensor types, including learning sensor resolution, fluid tank geometry, and fluid tank volume based on the output from the level sensor and an estimate of the amount of fluid delivered from the fluid tank.

[0006] One embodiment is one or more unique processes for self-learning of a fluid level sensor for sensor diagnostic purposes. In one embodiment, fluid level sensor sanity is determined by learning sensor resolution, fluid tank geometry, and fluid tank volume and making a sensor diagnostic determination based on the sensor resolution, fluid tank geometry, and fluid tank volume. Another embodiment is a unique system for self-learning of fluid level sensor sanity for a fluid storage tank.

[0007] Additional embodiments, forms, objects, features, advantages, aspects, and benefits shall become apparent from the description and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 is a schematic diagram of certain aspects of an example fluid storage system.

[0009] Figure 2 is a flow diagram depicting certain aspects of an example self-learning procedure for diagnostics of a fluid level sensor for a fluid storage system.

[0010] Figure 3 is a flow diagram depicting certain aspects of a sensor resolution learning circuit of a self-learning procedure for diagnostics of a fluid level sensor for Figure 2

[0011] Figure 4 is a flow diagram depicting certain aspects of a tank geometry learning circuit of a self-learning procedure for diagnostics of a fluid level sensor for Figure 2

[0012] Figure 5 is a flow diagram depicting certain aspects of a tank volume learning circuit of a self-learning procedure for diagnostics of a fluid level sensor for Figure 2 DETAILED DESCRIPTION

[0013] Referring to Figure 1 , an example system 100 for operating and managing a fluid storage system 120 for a vehicle 102 is shown. In Figure 1 ​​​In overview, system 100 includes an internal combustion engine 104 that operates with fuel to produce exhaust gas that is processed in an aftertreatment system 106 of vehicle 102. Fluid storage system 120 includes at least one tank such as a first tank 122 for storing fuel and a second tank 124 such as for storing dosing fluid. A first fluid such as fuel from first tank 122 can be supplied to engine 104 via fuel injectors 108. A second fluid such as diesel exhaust fluid (DEF) or other dosing fluid from second tank 124 can be supplied to the aftertreatment system via doser injectors 110. First tank 122 and second tank 124 each include a respective one of level sensors 126, 128 operable to measure a level therein.

[0014] In some embodiments, system 100 includes an electronic control unit (ECU) 130 connected to one or both of level sensors 126, 128 and to one or more of fuel injectors 108 and doser injectors 110. ECU 130 is also in operable communication with and configured to receive level readings from one or both of level sensors 126, 128. ECU 130 is also in operable communication with and configured to control operation of fuel injectors 108 to inject fuel from first tank 122 into combustion cylinders of engine 104 and / or doser injectors 110 to dose fluid from second tank 124 into aftertreatment system 106. ECU 130 is also in operable communication with and configured to determine and store an amount of fluid injected by fuel injectors 108 from first tank 122 and / or an amount of dosing fluid injected by doser injectors 110 from second tank 124.

[0015] ECU 130 is an example of a component of an electronic control system (ECS) configured and operable to execute operational logic defining various control, diagnostic, management, and / or regulatory functions. For example, a non-transitory memory medium can be configured with instructions executable by a processor to perform a number of actions, evaluations, or operations including those described herein. Operational logic of ECU 130 or other ECS components can be in the form of dedicated hardware such as a hardwired state machine, analog computer, programmed instructions, and / or different forms as can occur to those skilled in the art.

[0016] While the ECU 130 is depicted in the illustrated example as a single unit, it will be appreciated that one or more processors, one or more non-transitory memory media, and related components can be provided as multiple units or physical packages, or distributed or distributed among multiple units or physical packages. For example, one or more processors, such as programmable microprocessors or microcontrollers of a solid state integrated circuit type, can be provided in one or more control units, and can be implemented in any of several ways that combine or distribute control functions across the one or more control units in various ways. Other components or subsystems of the ECU 130 and / or its associated ECS can also be so configured or provided.

[0017] Referring now to the drawings Figure 2 An example program 200 that can be implemented and executed in whole or in part in conjunction with a system such as the system 100 and / or the ECU 130 is shown. The program 200 is one example of a method according to the present application for performing a diagnosis, such as a plausibility check or a functional anomaly indication, for a liquid level sensor 202, such as one or both of the liquid level sensors 126, 128 for a fluid tank 206, such as one or both of the first and second fluid tanks 122, 124.

[0018] The following illustrative flow description provides an example embodiment of performing a program for diagnosing a liquid level sensor 202. The operations shown are understood to be exemplary only, and operations can be combined or divided, added or removed, and reordered in whole or in part, unless specifically stated to the contrary herein. Certain operations shown can be implemented by a computer, such as the ECU 130, executing a computer program product on a non-transitory computer readable medium, where the computer program product includes instructions that cause the computer to perform one or more of the operations or issue commands to other devices to perform one or more of the operations.

[0019] The program 200 includes an operation 208 for receiving inputs that are used to learn properties of the fluid tank 206 that are used for diagnosis of the liquid level sensor 202. The inputs at operation 208 include liquid level measurements output from the liquid level sensor 202 and fluid consumption estimated from tracking an amount of fluid ejected from the fluid tank 206 by the ejector 204, such as the fluid ejector 108 or the doser ejector 110.

[0020] From operation 208, program 200 continues along parallel paths. A first path includes a default calibration circuit operation 210 that provides a default calibration for the performance of the liquid level sensor diagnostics. In one embodiment, the default calibration is for a worst case scenario for a combination of liquid level sensor resolution, tank volume, and tank geometry. In one embodiment, about 10 to 20 shifts or cycles of operating vehicle 100 are envisioned in order to provide sufficient learning to update the diagnostics based on learned conditions rather than default worst case conditions, as further discussed below.

[0021] A second path for program 200 in parallel with default calibration circuit operation 210 includes an operation 212 to process the inputs from operation 208 and an operation 214 involving adaptive circuits. Operation 212 can include, for example, filtering readings from liquid level sensor 202 and / or estimates of fluid consumption from injections or dosing provided by injectors 204. Operation 212 can also include correcting readings from liquid level sensor 202 and / or estimates of fluid consumption from injections or dosing provided by injectors 204. Operation 212 can also include removing outliers from the inputs or readings from liquid level sensor 202 and / or estimates of fluid consumption from injections or dosing provided by injectors 204 received from operation 208.

[0022] Once processing of the input data is complete at operation 212, program 200 continues operation using the input data at operation 214 involving data adaptive circuits. The adaptive circuits at operation 214 include a learning circuit operation 216, a learning confidence estimator operation 218, and a diagnostic margin circuit operation 220. Learning circuit operation 216 includes a sensor resolution learning circuit operation 222, a fluid tank geometry learning circuit operation 224, and a fluid tank volume learning circuit operation 226, embodiments of which are discussed below with reference to FIGS. 3-5. Figures 3 to 5 Further discussion.

[0023] Operation 218 applies learning confidence to quantify the uncertainty of the learned variables from learning circuit 216. In one embodiment, operation 218 applies bounds to the learned estimates of sensor resolution, tank geometry, and tank volume. Operation 220 applies a diagnostic margin, such as an error margin, to the learned variables based on the confidence estimates. In one embodiment, the margin for diagnosing a fault condition decreases as learning of the geometry of fuel tank 206 increases.

[0024] At an output operation 228, outputs from the default calibration circuit operation 210 and the adaptive circuit operation 214 are processed. The output operation 228 provides diagnostic outputs for the fluid level sensor 202 based on the default calibration circuit operation 210 and the adaptive circuit operation 214. The output operation 228 can include applying diagnostic fault thresholds 230 and / or diagnostic pass thresholds 232 to the fluid level sensor diagnostics. The fluid level sensor diagnostic outputs can include, for example, sensor plausibility checks, sensor stuck indication, diagnostics for high and / or low fluid level readings for sensor high and low plausibility checks, and / or sensor malfunction indication.

[0025] Alternatively or additionally, the fluid level sensor diagnostic outputs at operation 228 can include performing a sensor failure verification operation, and determining a sensor failure in further response to the sensor failure verification operation. In certain further embodiments, the sensor failure verification operation includes one or more operations such as averaging a number of sensor test results, incrementing a fault counter in response to sensor tests indicating a failed sensor, decrementing the fault counter in response to sensor tests indicating a passing sensor, integrating a fuel consumption estimate over a predetermined operational period and comparing the integrated value to fluid level sensor plausibility fault and / or pass thresholds, and modeling fluid storage in the fluid tank 206 and accounting for the storage in the sensor failure verification operation and / or plausibility check operation.

[0026] Referring to Figure 3 FIG. 3 shows an example procedure 300 that can be implemented and executed in whole or in part in conjunction with the procedure 200 and the sensor resolution learning circuit operation 222. The procedure 300 begins with an operation 302 to receive raw, pre-processed fluid level sensor readings from the fluid level sensor 202.

[0027] From the operation 302, the procedure 300 continues at an operation 304 to activate sensor resolution learning in response to one or more learning conditions being satisfied. For example, the learning activation at operation 304 can require that a tank level based enabling condition be satisfied to ensure that there is fluid in the fluid tank 206 that will enable a correct fluid level to be determined for sensor resolution. The learning activation at operation 304 can also require that one or more abort conditions not be satisfied. The abort conditions can include, for example, the engine 102 being off, an existing sensor fault, a frozen tank condition, presence of one or other fault conditions, etc.

[0028] At operation 308, the program 300 continues from operation 304 in response to the enable condition being satisfied and the abort condition not being satisfied. Operation 308 includes performing a plurality of measurement functions based on the level sensor readings received at operation 302. The measurement functions include an operation 310 to detect changes in sensor readings, an operation 312 to record the changes or increments in sensor readings and the associated one or more levels in the tank 206. The measurement functions also include an operation 314 to compare the changes in sensor readings and levels to a minimum number of required measurements needed or desired to determine sensor resolution.

[0029] The program 300 continues at operation 316 to process the measurements determined from operation 308. Operation 316 includes a first processing operation 318 to calculate an average value and a maximum value of the measurements from operation 308. Operation 316 includes a second processing operation 320 to reference a lookup table that cross-references sensor resolution numbers to average sensor reading changes and associated level changes. The program 300 continues at operation 322 to output a single sensor resolution number obtained from the lookup table based on the calculated average value and maximum value of the sensor reading changes and associated level changes.

[0030] Reference Figure 4 The program 400 is shown that can be implemented and executed in whole or in part in conjunction with the program 200 and tank geometry learning circuit operation 224. The program 400 includes an operation 402 to receive a plurality of inputs. A first input 404 includes the sensor resolution number from the output operation 322 of the sensor resolution learning circuit operation 222 of the program 300. A second input 406 includes the processed level sensor readings received from the input processing operation 212 of the program 200. A third input 408 includes the processed fluid consumption estimate from the input processing operation 212 of the program 200.

[0031] The program 400 continues at operation 410 to activate tank geometry learning in response to a learning condition being satisfied. For example, the learning activation at operation 412 can require that an enable condition based on tank levels be satisfied to ensure that there is a correct level of fluid in the fluid tank 206. The learning activation at operation 414 can also require that one or more abort conditions not be satisfied. In one embodiment, the operation 410 to activate learning can include the same or similar enable conditions and abort conditions as operation 304 of the program 300.

[0032] Program 400 continues at operation 416 in response to the fulfillment of an enable condition at operation 410 and the failure to meet one or more abort conditions. Operation 416 includes creating a bucket for containing input data from operation 410. In one embodiment, the bucket for input data is created for different liquid levels within fluid tank 206. For example, operation 416 includes operation 418 for looking up a reference for the bucket size to be used. Operation 416 also includes operation 420 for creating a non-volatile memory store for the input data in each bucket having a predetermined length. In an example embodiment, the memory store includes five arrays, each 100 elements long, with the capacity to support 100 buckets.

[0033] Program 400 continues at operation 422, which uses input data stored in the respective level tanks to begin learning the geometry of fuel tank 206. Operation 422 includes operation 424, which triggers the tank corresponding to the active level in fluid tank 206. At operation 426, during operation of engine 104, level readings from level sensor 202 and associated fluid consumption estimates are accumulated in the corresponding active level tank.

[0034] Operation 422 also includes operation 428 for repeatedly filling each active level tank with input data including level readings from level sensor 202 and associated fluid consumption estimates. In one embodiment, each active level tank is filled a maximum number of times. In one embodiment, the maximum number of times is 5. Operation 422 also includes operation 430 for providing each tank with a statistically aggregated volume for the learned volume of fluid consumption at each active level in fluid tank 206.

[0035] Program 400 continues at operation 432, which establishes a relationship between fluid consumption and level readings to learn the amount of fluid consumed by the fluid tank at or between various levels. This relationship may be, for example, an estimate of the cross-sectional area of ​​the fluid tank at each active level, an estimate of the volume of the fluid tank at or between each active level, the rate of change of the fluid tank at and / or between active levels, or the area or volume, etc.

[0036] refer to Figure 5 An example program 500 is shown that can be implemented and executed, either wholly or partially, in conjunction with program 200 and box volume learning circuit operation 226. Program 500 includes operation 502 for receiving the relationship between fluid consumption and liquid level readings created at operation 432 of program 400 during box geometry learning circuit operation 224.

[0037] The program 500 continues from the operation 502 at an operation 504 for activating tank volume learning in response to satisfying a learning condition. For example, the learning activation at operation 506 can require that a tank level based enable condition be satisfied to ensure that there is a correct level of fluid in the fluid tank 206. The learning activation at operation 508 can also require that one or more abort conditions not be satisfied. In one embodiment, the operation 504 for activating learning can include the same or similar enable conditions and abort conditions as the operation 304 of the program 300.

[0038] The program 500 continues from the operation 504 at an operation 510 for evaluating one or more conditions to be satisfied for calculating a volume of the fluid tank 206. A first condition 512 includes a total number of learned fluid consumption amounts for the fluid tank 206 being greater than a threshold value. A second condition 514 is a total number of active level buckets having at least a minimum number of learned consumption values being greater than a minimum threshold bucket number. In one embodiment, the minimum bucket number is at least three buckets having the minimum number of required learned fluid consumption values.

[0039] A third condition 516 is a difference between a highest level bucket having at least a minimum number of learned fluid consumption values and a lowest level bucket having at least a minimum number of learned consumption values being greater than a threshold value. For example, to learn a tank volume, the learned fluid consumption values should not come only from adjacent active levels of the fluid tank 206, but include at least one active level at or near a top of the fuel tank 206 (full or near full condition) and at least one active level at or near a bottom of the fuel tank 206 (empty or near empty condition).

[0040] In response to satisfying the conditions at the operation 510, the program 500 continues from the operation 510 at an operation 518. The operation 518 includes providing an output corresponding to a calculation of a total volume of the fluid tank 206 based on the learned fluid consumption values at the various levels of the fluid tank 206.

[0041] Based on the output of the diagnosis when a functional anomaly of the sensor is successfully detected, one or more of the following actions, etc., can be taken. One action can include shutting down fluid injection in extreme cases upon determining a safe harbor condition. Another action includes reducing fluid injection amount using a transient maximum limit or a percentage based derate. Another action includes setting the level value to a default value. Another action includes illuminating a functional anomaly light on a vehicle dashboard. Another action includes wirelessly sending a message to an operator or fleet management service. Another action includes changing engine operation calibration to produce less NOx emissions.

[0042] The systems and procedures disclosed herein can be used with any type of level sensor and fluid tank size / geometry. No software redesign or calibration changes are required for different fluid tank constructions or applications because the sensor resolution, fluid tank geometry, and fluid tank volume are learned and adaptable to any fluid tank construction or geometry. Furthermore, since the diagnostic procedures are based on fluid consumption, in addition to diagnosing “stuck” or malfunctioning sensors, diagnostics for high sensor output reasonableness (overly high readings for active levels) and low sensor output reasonableness (underly low readings for active levels) are also possible.

[0043] It should be understood that terms such as “non-transitory memory,” “one or more non-transitory memory media,” and “non-transitory memory device” refer to several types of devices and storage media that can be configured to store information (such as data or instructions) that can be read or executed by a processor or other components of a computer system, and such terms include and cover a single or single device or medium storing such information, multiple devices or media on or in which a corresponding portion of such information is stored, and multiple devices or media on or in which multiple copies of such information are stored.

[0044] It should be understood that terms such as “determine,” “determined,” and “determining,” when used in conjunction with control methods or processes, electronic control systems or controllers, electronic control, or components or operations of the foregoing, encompass several actions, configurations, devices, operations, and techniques, including but not limited to the calculation or operation of parameters or values; obtaining parameters or values ​​from lookup tables or using lookup operations; receiving parameters or values ​​from data links or network communications; and receiving electronic signals indicating parameters or values. For example (Voltage, frequency, current, or pulse width modulation (PWM) signal); receive sensor output indicating parameters or values; receive other outputs or inputs indicating parameters or values; read parameters or values ​​from memory locations on computer-readable media; receive parameters or values ​​as runtime parameters; and / or by receiving parameters or values ​​that can be interpreted as parameters through calculation and / or by referencing default values ​​interpreted as parameter values.

[0045] As shown in this detailed description, the present invention contemplates numerous embodiments, several examples of which will now be further illustrated. A first example embodiment is a diagnostic method for a liquid level sensor in a fluid tank of a vehicle, the method comprising: measuring a plurality of liquid levels in the fluid tank with the liquid level sensor; determining a fluid consumption associated with the plurality of liquid level measurements; learning a sensor resolution of the liquid level sensor, a geometry of the fluid tank, and a volume of the fluid tank based on the plurality of liquid level measurements and the fluid consumption; and outputting a diagnosis for the liquid level sensor based on the learned sensor resolution, the learned geometry of the fluid tank, and the learned volume of the fluid tank.

[0046] In one embodiment, the method comprises outputting a diagnosis for the liquid level sensor based on assumed values for the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank while learning the sensor resolution of the liquid level sensor, the geometry of the fluid tank, and the volume of the fluid tank.

[0047] In one embodiment, the learning comprises first learning the sensor resolution of the liquid level sensor, then learning the geometry of the fluid tank, and then learning the volume of the fluid tank.

[0048] In one embodiment, the method comprises estimating a confidence metric for the liquid level sensor diagnosis based on the learning of the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank.

[0049] In another embodiment, the method comprises assigning a diagnostic margin based on the confidence metric; and determining a threshold for passing and failing the liquid level sensor diagnosis based on the diagnostic margin.

[0050] In one embodiment, the fluid tank contains diesel exhaust fluid for dosing into an aftertreatment system.

[0051] In one embodiment, the learning the sensor resolution comprises: detecting an output change from the liquid level sensor; recording the output change from the liquid level sensor and a change in liquid level in the fluid tank associated with the output change from the liquid level sensor; and determining the sensor resolution based on the output change from the liquid level sensor and the change in liquid level in the fluid tank associated with the output change from the liquid level sensor.

[0052] In another embodiment, the sensor resolution is determined based on an average of output changes from the liquid level sensor and associated changes in liquid level in the fluid tank and a maximum output change from the liquid level sensor and an associated change in liquid level in the fluid tank.

[0053] In another embodiment, learning the tank geometry is based on: a number of sensor resolutions, a measured level in the fluid tank, and a fluid consumption, and determining a relationship between the level in the fluid tank and the fluid consumption associated with the level in the fluid tank.

[0054] In yet another embodiment, the learning tank geometry action includes learning a relationship for fluid consumption at each of a plurality of levels in the fluid tank.

[0055] In yet another embodiment, learning the tank volume is based on: determining a relationship between fluid consumption and a plurality of levels in the fluid tank a plurality of times for each of the plurality of levels in the fluid tank.

[0056] In yet another embodiment, the learning tank volume requires that a difference between a highest level having a plurality of learned relationships between fluid consumption and a level in the fluid tank and a lowest level having a plurality of learned relationships between fluid consumption and a level in the fluid tank is greater than a threshold value.

[0057] According to another aspect of the present disclosure, there is provided a system for diagnosing a level sensor in a fluid tank of a vehicle. The system includes an electronic control system including at least one electronic control unit configured to execute instructions stored in one or more non-transitory memory media to: measure a plurality of levels in the fluid tank with a fluid sensor; determine a fluid consumption associated with the measured values of the plurality of levels in the fluid tank; learn a sensor resolution of the level sensor, learn a geometry of the fluid tank, and learn a volume of the fluid tank based on the plurality of level measurements and the fluid consumption; and output a diagnosis for the level sensor based on the learned sensor resolution, the learned geometry of the fluid tank, and the learned volume of the fluid tank.

[0058] In one embodiment, the electronic control system is configured to output the diagnosis for the level sensor based on assumed values for the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank while learning the sensor resolution of the level sensor, the geometry of the fluid tank, and the volume of the fluid tank.

[0059] In another embodiment, the electronic control system is configured to first learn the sensor resolution of the level sensor, then learn the geometry of the fluid tank, and then learn the volume of the fluid tank.

[0060] In another embodiment, the electronic control system is configured to estimate a confidence metric for the level sensor diagnosis based on the learning of the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank.

[0061] In yet another embodiment, the electronic control system is configured to assign a diagnostic margin based on the confidence metric; and determine a threshold for passing and failing the level sensor diagnosis based on the diagnostic margin.

[0062] In one embodiment, the electronic control system is configured to learn the sensor resolution by: detecting an output change from the level sensor; recording the output change from the level sensor and a change in the level in the fluid tank associated with the output change from the level sensor; and determining the sensor resolution based on the output change from the level sensor and the change in the level in the fluid tank associated with the output change from the level sensor.

[0063] In another embodiment, the electronic control system is configured to learn the tank geometry based on: the number of sensor resolutions, the measured level in the fluid tank, and the amount of fluid consumed, and determine a relationship between the level in the fluid tank and the amount of fluid consumed associated with the level in the fluid tank.

[0064] In yet another embodiment, the electronic control system is configured to learn the tank volume based on: determining the relationship between the amount of fluid consumed and the plurality of levels in the fluid tank a plurality of times for each of the plurality of levels in the fluid tank.

[0065] While example embodiments of the application have been illustrated and described in detail in the drawings and foregoing description, the same should be considered as illustrative and not restrictive in character, it being understood that only certain embodiments have been shown and described and that all changes and modifications that come within the spirit of the application as claimed below are desired to be protected. It is understood that while the use of words such as "preferably," "preferably," "preferred," or "more preferred" utilized in the description above indicates that the feature so described can be more desirable, it is understood that such features are not necessary and embodiments lacking in such features can be considered within the scope of the application, the scope of the application being defined in part by the claims below. It is intended that when read in the claims, the use of the word "a" or "an" to refer to an element or feature of the application indicates that the claims are not limited to one element or feature but rather potentially more than one element or feature. When the language "at least a portion" and / or "a portion" is used in the language of the specification, unless specifically stated to the contrary, the item can comprise a portion and / or the whole.

Claims

1. A diagnostic method for a liquid level sensor in a fluid tank of a vehicle, the method comprising: The liquid level sensor is used to measure multiple liquid levels in the fluid tank; Determine the fluid consumption associated with the measured values ​​of the plurality of liquid levels; The sensor resolution of the liquid level sensor, the geometry of the fluid tank, and the volume of the fluid tank are learned based on the multiple liquid level measurements and the fluid consumption. as well as The diagnostics for the level sensor are output based on the learned sensor resolution, the learned geometry of the fluid tank, and the learned volume of the fluid tank.

2. The method of claim 1, further comprising, while learning the sensor resolution of the level sensor, the geometry of the fluid tank, and the volume of the fluid tank, outputting the diagnosis for the level sensor based on assumed values ​​for the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank.

3. The method according to claim 1, wherein, The learning process includes first learning the sensor resolution of the liquid level sensor, then learning the geometry of the fluid tank, and then learning the volume of the fluid tank.

4. The method of claim 1, further comprising estimating a confidence metric for the level sensor diagnostic based on the learning of the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank.

5. The method according to claim 4, further comprising: Allocate diagnostic margins based on the aforementioned confidence metric; as well as The threshold for diagnosing pass and fail of the level sensor is determined based on the diagnostic margin.

6. The method according to claim 1, wherein, The fluid tank contains diesel exhaust treatment fluid for metering feed into the aftertreatment system.

7. The method according to claim 1, wherein, Learning the sensor resolution includes: Detect changes in the output from the liquid level sensor; Record the output change from the level sensor and the level change in the fluid tank associated with the output change from the level sensor; and The sensor resolution is determined based on the output change from the level sensor and the level change in the fluid tank that is associated with the output change from the level sensor.

8. The method according to claim 7, wherein, The sensor resolution is determined based on the average of the output change from the level sensor and the associated level change in the fluid tank, as well as the maximum output change from the level sensor and the associated level change in the fluid tank.

9. The method according to claim 7, wherein, The geometry of the chamber is learned based on: the sensor resolution, the measured liquid level in the fluid chamber, and the fluid consumption, and the relationship between the liquid level in the fluid chamber and the fluid consumption associated with the liquid level in the fluid chamber is determined.

10. The method according to claim 9, wherein, Learning the box geometry involves learning the relationship between fluid consumption at each of the multiple liquid levels in the fluid box.

11. The method according to claim 9, wherein, The learning of the tank volume is based on: for each of the plurality of liquid levels in the fluid tank, repeatedly determining the relationship between the fluid consumption and the plurality of liquid levels in the fluid tank.

12. The method according to claim 11, wherein, The learning of the tank volume requires that the difference between the highest liquid level having multiple learned relationships between fluid consumption and the liquid level in the fluid tank and the lowest liquid level having multiple learned relationships between fluid consumption and the liquid level in the fluid tank is greater than a threshold.

13. A system for diagnosing a liquid level sensor in a fluid tank of a vehicle, the system comprising: An electronic control system, comprising at least one electronic control unit configured to execute instructions stored in one or more non-transitory memory media to: Multiple liquid levels in the fluid tank are measured using fluid sensors; Determine the fluid consumption associated with the measured values ​​of the plurality of liquid levels in the fluid tank; The sensor resolution of the liquid level sensor, the geometry of the fluid tank, and the volume of the fluid tank are learned based on the multiple liquid level measurements and the fluid consumption. as well as The diagnostics for the level sensor are output based on the learned sensor resolution, the learned geometry of the fluid tank, and the learned volume of the fluid tank.

14. The system according to claim 13, wherein, The electronic control system is configured as follows: While learning the sensor resolution of the liquid level sensor, the geometry of the fluid tank, and the volume of the fluid tank, the diagnostics for the liquid level sensor are output based on assumed values ​​for the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank.

15. The system according to claim 14, wherein, The electronic control system is configured to first learn the sensor resolution of the liquid level sensor, then learn the geometry of the fluid tank, and then learn the volume of the fluid tank.

16. The system according to claim 14, wherein, The electronic control system is configured to estimate a confidence metric for the level sensor diagnostics based on the learning of the sensor resolution, the geometry of the fluid tank, and the volume of the fluid tank.

17. The system according to claim 16, wherein, The electronic control system is configured as follows: Allocate diagnostic margins based on the aforementioned confidence metric; and The threshold for diagnosing pass and fail of the level sensor is determined based on the diagnostic margin.

18. The system according to claim 13, wherein, The electronic control system is configured to learn the sensor resolution by: Detect changes in the output from the liquid level sensor; Record the output change from the level sensor and the level change in the fluid tank associated with the output change from the level sensor; and The sensor resolution is determined based on the output change from the level sensor and the level change in the fluid tank that is associated with the output change from the level sensor.

19. The system according to claim 18, wherein, The electronic control system is configured to learn the box geometry based on: the number of sensor resolutions, the measured liquid level in the fluid box, and the fluid consumption, and to determine the relationship between the liquid level in the fluid box and the fluid consumption associated with the liquid level in the fluid box.

20. The system according to claim 19, wherein, The electronic control system is configured to learn the tank volume based on the following: for each of the plurality of liquid levels in the fluid tank, repeatedly determining the relationship between the fluid consumption and the plurality of liquid levels in the fluid tank.

Images