Method for diagnosing operation of active air flow regulation system

Abstract

The present invention relates to an active air flow regulation system and a method for diagnosing the operation of an active air flow regulation system of a vehicle, the active air flow regulation system comprising: a blocking mechanical member; -an electromechanical system provided with a housing accommodating a magnetic motor driven by the control circuit and coupled to a movement conversion device for moving the blocking mechanical member; -an electronic communication line for exchanging information with or receiving commands from a DCU (8) of the vehicle, characterized in that the method comprises a series of steps comprising:-during an operational cycle of the active air flow regulation system, storing in the memory a sequence of values S of data, the data is acquired by a control circuit in the electromechanical system; applying an algorithm statistical analysis model to the sequence of values of data stored in the memory in order to determine singular points; -identifying an accidental error associated with the singular point in order to provide information about the state of the active air flow regulation system.

Description

Technical Field

[0001] The invention relates to the field of active air flow regulation of vehicles and in particular to the modification of the air permeability coefficient of the vehicle by the use of motorized fins or spoilers, the thermal control of the vehicle, in particular the engine, the brake system and more generally any component whose temperature is actively regulated, in particular motor vehicles with internal combustion engines or electric motors or hybrid motors, by means of an active air flow regulation system to drive a blocking member such as a motorized air shutter on a grille or curtain. The blocking member is designed to vary the amount of air passing through the radiator and engine compartment of the vehicle and to adjust the air flow in order to manage the efficiency of the vehicle's engine and various main units (combustion engine, electric traction motor and its inverter, gearbox, battery pack, etc.). With the development of anti-pollution standards, engine temperature management has become a major area of ​​development in engine design. Engine cooling is now optimized to control the temperature of various engine zones while limiting energy consumption.

[0002] The purpose of the thermal management circuit is to maintain the various internal components of the engine (cylinders, cylinder head, etc.) and, where applicable, peripheral components (eg, turbocharger) at their ideal operating temperatures.

[0003] If the engine temperature is too high, the gases inside the cylinder are hotter, which means there is less gas in the cylinder and there is a greater risk of auto-ignition (and knocking). On the other hand, when components deform more, there is a greater risk of breakage.

[0004] On the other hand, if the engine is cold, the friction between the piston and the cylinder liner is generally higher and the combustion is incomplete, resulting in significantly higher pollutant emissions.

[0005] On many recent vehicles, the fins of the grille located in front of the radiator are hinged and motorized to control the air flow through the radiator. In practice, when the vehicle is traveling at high speed and low load, high air flow is not necessary to properly cool the engine.

[0006] Furthermore, the air flowing through the radiator disturbs the main flow around the vehicle, thereby generating aerodynamic drag. It may also be advantageous to block the air inlet upstream of the radiator with a shutter. In some cases, the shutter has only two positions (a fully open position or a fully closed position).

[0007] Unfortunately, the CO2 reduction benefit of these active shutter calenders is lost if the shutters remain locked in the open or closed position, or in an intermediate position, or if the shutters suffer damage, for example in the event of an impact, or if one or more shutters are missing. The correct operation of the active shutter grilles has an impact on the pollutant emissions of the vehicle. It is therefore crucial, and even legally mandatory, to ensure the correct operation of the air shutter device and to inform the driver of any deterioration in the vehicle's fuel consumption and pollution levels every time the vehicle is driven, thereby prompting him or her to quickly go to the garage to restore the vehicle's legal emission levels. To this end, solutions have been provided for checking the general functionality of the air shutter device by means of an on-board diagnostic system. For electric vehicles, thermal control of the motor and battery is also an important matter, with the same requirements as aerodynamic efficiency: a faulty active air flow regulation system may be detrimental to the vehicle's air penetration and therefore have a negative impact on the vehicle's mileage. Prior art

[0008] Various solutions are known in the prior art to meet this purpose of detecting a fault in an active shutter of a grid.

[0009] Patent DE102018108162 describes an actuator control method and an actuator in which a DC motor drive system moves a first body relative to a second body, the first body sliding along a contact surface of the second body. The principle of the solution is based on current analysis for evaluating the behavior of the actuator by providing an area with a changing mechanical resistance on the contact surface between the roller gate and the guide rail.

[0010] This solution has several disadvantages.

[0011] Firstly, it requires the adjustment of a specific reference zone on the system, which is necessary for detecting anomalies using this prior art solution. Anomaly detection is strictly limited to detecting anomalies in this reference zone and never allows detecting anomalies in general in mechanical and pneumatic transmission chains. This reference zone may also wear out, rendering the signal evaluation invalid.

[0012] Secondly, it is limited to electromechanical systems using DC motors and cannot be implemented for systems using stepper motors; the use of torque (or DC current) measurement imposes closed-loop motor control, which is incompatible with stepper motor control, making prior art solutions unusable or difficult in this scenario.

[0013] Thirdly, the signal analyzed is the current reflecting the torque transmitted to the computer. However, the update frequency of this information is of the order of 10 μs, while the transmission capacity of the multiplexed LIN type protocol is of the order of 100 ms, resulting in incompatibility and insufficient bandwidth to transmit the necessary information. The transmission of all the information acquired is not suitable for the very limited bandwidth of the communication network used for the vehicle control electronics.

[0014] Patent US2020300156 describes another abnormality detection system for an engine cooling water recirculation system, which includes: a grille gate capable of regulating the circulating air flow rate from the outside of the vehicle to the environment of the engine body, wherein four learning neural networks are stored, and the four learning neural networks are obtained by using at least five parameters (including the engine cooling water temperature at engine start, the amount of air sucked into the engine, the amount of fuel injected into the engine, the outside air temperature and the vehicle speed) as input parameters of the neural network and using the measured value of the engine cooling water temperature as training data to learn the weights of four states, and the four states include a state in which the grille gate is closed and the air blown by the blower does not circulate through the air conditioning heater, a state in which the grille gate is open and the air blown by the blower does not circulate through the air conditioning heater, a state in which the grille gate is closed and the air blown by the blower circulates through the air conditioning heater, and a state in which the grille gate is open and the air blown by the blower circulates through the air conditioning heater. The engine cooling water temperature is estimated from the five parameters using any of the four learning neural networks corresponding to the current state of the grille shutter and the circulation state of the air blown by the blower in the air conditioning heater. An engine cooling water recirculation system fault is detected based on the estimated engine cooling water temperature value.

[0015] The drawback of this solution is that it provides information related to the prediction of the temperature of the cooling water, which does not allow us to characterize anomalies in the kinematics and pneumatic chain of the shut-off system.

[0016] Moreover, this solution requires the deployment of four neural networks and thus involves extremely complex processing.

[0017] Patent DE10201201469 describes an air brake device also having an adjustment drive device, which is connected to at least one air brake in a moving and force-transmitting manner so as to adjust the at least one air brake between its predetermined operating positions. The air brake device also has a control device coupled to the adjustment drive device in a signal transmission manner, a control device connected to the control device, a storage device coupled in a signal transmission manner, and a data acquisition device that captures operating data from the air brake device memory device, wherein the differences in operating positions representing a set of reference operating data are stored in the memory device.

[0018] Patent US213338870 describes another system solution and method for diagnosing a mechanical fault affecting an active grille shutter (AGS) system of an automobile. The method includes receiving at least one mechanical fault status signal (410) indicating that a specific mechanical fault condition has been detected in the AGS system. The method also involves receiving a temperature signal (418) indicating a temperature at a location near a grille shutter of the AGS system, and determining whether to perform an extended diagnosis based at least on the mechanical fault status signal and the temperature signal.

[0019] Disadvantages of existing technology

[0020] Prior art solutions cannot reliably characterize anomalies in the state of the mechanical and pneumatic shut-off chain without any mechanical modification of the chain, such as deformation of the shutter or partial breakage of the air inlet system.

[0021] The solution provided by the present invention

[0022] In order to overcome these disadvantages and to provide a diagnostic function for the entire system without modifying the blocking member, data internal to the electromechanical system that activates the blocking member or, where applicable, generated by the blocking member are used to characterize air flow faults. To this end, the invention in its most general sense relates to a method for diagnosing the operation of an active air flow control system of a vehicle having the features set forth in claim 1 and to a system having the features set forth in the system claims.

[0023] The method includes:

[0024] Blocking of mechanical components;

[0025] an electromechanical system provided with a housing housing a magnetic motor driven by a control circuit and coupled to movement conversion means for moving the blocking mechanical member;

[0026] an electronic communication line for exchanging information with the vehicle's DCU or receiving commands from the vehicle's DCU,

[0027] The method comprises a series of steps, the series of steps comprising:

[0028] storing in a memory a sequence S of values ​​of data acquired by a control circuit inside the electromechanical system during an operating cycle of the active air flow control system;

[0029] Applying the algorithmic statistical analysis model to the numerical sequence S of data stored in the memory,

[0030] In order to determine the singularity;

[0031] Identify sporadic errors associated with singularities to provide information about the status of the active air flow conditioning system.

[0032] In a variant, the sporadic error identification step is performed by a control circuit which transmits information representative of the identified sporadic error to the DCU.

[0033] In another variation, the digital data sequence stored in the memory is:

[0034] - pre-processing at a first processing frequency by the control circuit; and

[0035] - post-processing by the DCU / ECU of the vehicle at a second processing frequency lower than the first processing frequency to complete the analysis and identify sporadic errors;

[0036] The data set preprocessed by the control circuit is transmitted to the DCU / ECU via a command line.

[0037] In one embodiment, the digital data sequence S stored in the memory comprises:

[0038] - data of the first type; and

[0039] - At least a second type of data different from the first type.

[0040] In another embodiment, the method comprises the following steps as additional steps:

[0041] - Integrate sporadic errors according to the instructions transmitted by the vehicle's DCU, thereby generating an integrated diagnostic status;

[0042] -Transmits the integrated diagnostic status to the vehicle's DCU.

[0043] For the purposes of this patent, "ECU" means an "electronic control unit" that controls specific vehicle functions.

[0044] However, driving safety, enhanced comfort and infotainment functions lead to unsustainable system complexity - up to hundreds of ECUs per vehicle.

[0045] It is also known to use a domain controller (DCU), the purpose of which is to centralize via this DCU (domain controller) instead of using multiple ECUs. The rise of automotive domain controllers is particularly driven by vehicle automation, which requires combining data from different sensors, high-speed processing, and must meet the most stringent safety requirements. It must also provide the possibility of continuous development of new features and functions via over-the-air (OTA) updates throughout the life of the vehicle.

[0046] The present invention will refer indiscriminately to an ECU or a DCU, as the nature of the controller depends on the choice of computer architecture of the vehicle in which the invention is implemented.

[0047] Additionally, the statistical analysis algorithm may provide a supervised learning sequence, an unsupervised learning sequence, or a reinforcement learning sequence.

[0048] Alternatively, the statistical analysis algorithm may be equipped with a cluster selection method.

[0049] In a variant, the blocking mechanism has no dedicated diagnostic elements external to the electromechanical system.

[0050] In the allowed variations, the digital data s i Relevant to the operation of magnet motors of electromechanical systems.

[0051] In the expected alternative, the digital data s i It is the motor phase current measurement.

[0052] In this case, the digital data s i It can be a measure of the rotor load angle of a magnetic motor.

[0053] In one embodiment, the algorithmic model for statistical analysis to determine singularities provides for a comparison between a sequence S of stored sampled digital data and a reference data map.

[0054] In a second embodiment, the algorithmic model for statistical analysis to determine singularities provides for the detection of singularities in a sequence S of stored sampled digital data.

[0055] In a third variant, the algorithmic model used for the statistical analysis to determine the singularities consists in subjecting the sequence S of stored sampled digital data to a model obtained by training a neural network according to training data corresponding to the nominal operation of the blocking mechanical member.

[0056] In all cases, a reference sequence S of sampled digital data can be stored for each electromechanical system at the end of the assembly line.r .

[0057] In this case, the algorithmic model for statistical analysis can compare a sequence of sampled digital data with a reference sequence of sampled digital data in order to detect behavioral drifts that could cause degradation of the electromechanical system or to verify normal wear behavior of the system.

[0058] In one embodiment, the stored sequence S of sampled digital data is derived from digital data items that depend on mechanical loads in the active mesh system.

[0059] In this case, identifying singular points in the sequence S of stored sampled digital data can be achieved by knowing the following:

[0060] The mass or inertia to be moved in the active grid system;

[0061] and / or mechanical friction in the system;

[0062] and / or pneumatic loads imposed on the system;

[0063] and / or system temperature.

[0064] For the purposes of this patent, a "singularity" is defined as a point or a subset of points in a stored data sequence that differs from a reference sequence corresponding to nominal operation, by comparison with one or more pre-recorded nominal data sequences, or by the difference between the signature of the data sequence and the signature of the nominal data sequence acquired during a preliminary learning phase (obtained by the same signature calculation function), or as a singularity in the Gaussian statistical sense, or a point or a set of points that differs from the projection of the median curve formed by the previous points, which point or set of points significantly differs from the average value of the fluctuations of the stored points relative to the median curve, significant being understood as "exceeding at least 20% of the amplitude of fluctuations in normal or nominal operation".

[0065] In particular, a distinction is made between data for which a signature is sought (torque, current, position, etc.) and information (such as temperature) stored in an internal memory of the mechatronic system, in particular in a volatile memory or a non-volatile memory of a controller.

[0066] In a variant, the algorithmic model used for the statistical analysis to determine the singularities consists in subjecting the sequence S of stored sampled digital data to a model obtained by training a neural network according to training data corresponding to the nominal operation of the blocking mechanical member.

[0067] In certain embodiments, a reference sequence of sampled digital data is stored for each electromechanical system at the end of an assembly line.

[0068] Advantageously, the algorithmic model for statistical analysis is capable of comparing a sequence of sampled digital data with a reference sequence of sampled digital data in order to detect behavioral drift.

[0069] In one embodiment, the stored sequence S of sampled digital data is derived from digital data items that depend on mechanical loads in the active mesh system.

[0070] According to a particular embodiment, identifying a singularity in a sequence S of stored sampled digital data is accomplished by knowing:

[0071] The mass or inertia to be moved in the active grid system;

[0072] and / or mechanical friction in the system;

[0073] and / or pneumatic loads imposed on the system;

[0074] The invention also relates to an electromechanical active air flow control system (1) for a vehicle, the electromechanical active air flow control system comprising:

[0075] - blocking of mechanical components;

[0076] - an electromechanical system provided with a housing housing a magnetic motor (10) driven by a control circuit and coupled to movement conversion means for moving the blocking mechanical member;

[0077] - an electronic communication line for exchanging information with the DCU of the vehicle or receiving commands from the DCU of the vehicle,

[0078] The method comprises a microcontroller or a microprocessor executing a computer program stored in its read-only memory to control a series of steps, the series of steps comprising:

[0079] - storing in the memory a sequence S of sampled digital data during an operating cycle of the active air flow control system, wherein the sequence of sampled digital data consists of a set of digital data s i The digital data set is obtained by the control circuit in the electromechanical system (3);

[0080] - applying an algorithmic statistical analysis model to said sequence S of sampled digital data stored in a memory in order to determine singularities;

[0081] - sporadic errors associated with singularities are detected by the control circuit and transmitted to the DCU, which sporadic errors are information about the state of the active air flow control system (1). DETAILED DESCRIPTION

[0082] The invention will be better understood by reading the following description of different non-limiting embodiments with reference to the accompanying drawings, in which:

[0083] Figure 1 FIG. 1 shows a schematic diagram of an active air flow regulation system driven by a stepper motor according to the present invention,

[0084] Figure 2 A schematic diagram showing a variation of an active air flow control system driven by a sensor BLDC motor according to the present invention is shown,

[0085] Figure 3 A schematic diagram showing a variation of an active air flow control system driven by a sensorless BLDC motor according to the present invention is shown,

[0086] Figure 4 shows an example of a diagnostic algorithm for fault identification specifically for a gate system,

[0087] Figure 5 Shown based on Figure 4 An example of a learning algorithm for fault identification is given in the algorithm given in

[0088] Figure 6 An exploded view of the active damper air conditioning system is shown,

[0089] Figure 7 shows an exploded view of an active curtain air conditioning system,

[0090] Figures 8 to 12 Shown with Figure 6 Various examples of system-related failures are shown,

[0091] Figures 13 to 16 Shown with Figure 7 Various examples of system-related failures are shown,

[0092] ] FIG. 17 (17a, 17b, 17c, 17d) shows a Fresnel diagram of a three-phase motor, allowing the load angle of the motor to be visualized,

[0093] Fig.18 A method for identifying a singularity by acquiring a series of data based on measurements of the power consumed by the motor is shown,

[0094] Fig.19 shows that can be used as Figure 4 An example of a supervised learning algorithm that underlies the decisions of the algorithm presented in .

[0095] General principles of active air flow control systems

[0096] Figure 1The present invention shows an overall view of an active air flow regulation system (1). The regulation system is constituted by a mechanical blocking member (2) positioned on the grille of the vehicle and capable of blocking the air flow to the radiator, and comprises an electromechanical system (3) controlling the positioning of the mechanical blocking member (2), such as a shutter or a roller, so as to allow the air flow received by the radiator to be adjusted according to the state of the mechanical blocking member (2).

[0097] The object of the invention is to automate the detection and characterization of anomalies in mechanical and / or pneumatic air flow control elements, in particular the disappearance of a flap, or a change in flap geometry, for example caused by an impact, even if these anomalies do not change the overall kinematics of the electromechanical air flow control system. Of course, the invention also enables the detection of faults in an electromechanical air flow control system, such as blockages or jams.

[0098] The electromechanical system (3) comprises a motion transmission chain (20) driven by a multi-phase permanent magnet synchronous motor, hereinafter referred to as a magnetic motor (10), which is controlled by an electronic control circuit (30), for example Figure 1 Three-phase motor shown.

[0099] The invention consists in utilizing the signals available on the electronic control circuit (30) of the electromechanical system (3), without the need to add one or more additional sensors on the blocking mechanical member (2) of the active air flow control system (1), so that the electromechanical system (3) according to the invention, when integrated into an active air flow control system (1) that initially lacks the function of providing diagnostic information, can provide this function. However, when additional sensors are already implemented in the active air flow control system (1), for example on the mechanical blocking member (2), the invention can also use the signals from these sensors, as long as these signals are processed by the electronic control circuit (30).

[0100] Figure 1 The illustrated situation shows a configuration in which the magnetic motor 10 is step-driven. The electromechanical system (3) is equipped with:

[0101] an electrical connector (6) which receives an electrical power supply (5) from the vehicle's DC power network;

[0102] an electrical communication line (7) comprising commands from a domain control unit (DCU) (8) of the vehicle and information generated by the electromechanical system (3) itself and intended for the DCU (8) of said vehicle;

[0103] Electrical protection and filtering modules (9) that comply with applicable electrical and electromagnetic standards;

[0104] A control circuit (11) such as a microcontroller provided with a memory, which generates vehicle communications, control functions and drives the magnetic motor (10) by association with a transistor drive module;

[0105] A magnetic motor (10) that produces the desired displacement and torque;

[0106] a multi-phase inverter (12), preferably two-phase or three-phase, comprising MOSFET transistors that supply power to the magnetic motor (10);

[0107] A kinematic transmission chain (20) for reducing or converting the motion of the rotor of the magnetic motor (10), the input stage of which is driven by the rotor of the magnetic motor (10) and the output stage (21) of which is coupled to the blocking mechanical member (2).

[0108] It is important to note that the control circuit (11), the transistor driver module and the multiphase inverter (12) can be integrated into a specific component (13) known to the layman by the acronym SOC (System on a Chip). Preferably, the data sequence is generated by combining at least two different types of information, which can be derived from the following elements without limitation:

[0109] one (or more) vibration sensor (13) measuring the vibration level of the moving subassembly in the context of the active air flow conditioning system (1);

[0110] one (or more) temperature sensor (14) which measures the internal temperature of the electromechanical system (3) and enables the local temperature of the area of ​​the active air flow conditioning system (1) to be deduced;

[0111] one (or more) position sensor (15) which measures the continuous position of the rotor of the magnetic motor (10) or of a movable element in the electromechanical system (3) and enables reconstruction of the load angle of the magnetic motor (10);

[0112] One (or more) electrical sensors (16) measuring one or more electrical parameters, such as voltage or current, at the multiphase inverter (12) and, when recombining the one or more electrical parameters, enabling reconstruction of the load angle of the magnetic motor (10).

[0113] Combining information from several sensor types can be advantageous when there is strong economic pressure on the components integrated in the actuator (e.g. very basic control electronics or low-sensitivity sensors that do not allow to extract a singularity from the information from a single type of sensor). However, the singularity can be isolated by the cooperation of weak signals and then identified in the sequence of sampled data resulting from the aggregation of data from sources of different types. However, this approach is not limiting, since once the economic pressure is eliminated, more complex systems can be envisaged and sporadic errors can be extracted from a sequence of data presenting a single type of information.

[0114] exist Figure 1 In the example shown, a typical scenario can be described as follows. The control circuit (11) receives movement instructions from the DCU (8) and then generates an appropriate drive sequence for the switches of the multiphase inverter (12) in order to produce the movement. The movement instruction from the DCU (8) can be a complete or partial opening or closing movement, or an absolute position of the stroke to be achieved. The invention consists in acquiring a data sequence by the control circuit (11), storing the data sequence in a memory, analyzing the data sequence using digital processing capable of detecting singularities, and transmitting the occasional errors associated with the singularities to the DCU (8) via the communication line (7).

[0115] The data sequence can be acquired for any movement, or intermittently according to instructions from the DCU (8) or the control circuit (11).

[0116] The control circuit (11) is able to extract singularities from the sequence of stored data using statistical processing of varying complexity, depending on the functional requirements arising from the type of fault sought and the desired sensitivity. For example, a fault caused by a sudden degradation, such as a gate breakage blocking the stroke of the blocking mechanical member (2), is easier to detect than a gradual aging of the actuator, which results in a slight degradation of performance close to the minimum requirements. Therefore, such a singularity may be a symptom of various faults and it is necessary to apply appropriate processing algorithms to generate from this error an occasional error for transmission to the DCU (8). "Occasional error" refers to an error code that is understandable to the DCU (8) and may imply various faults encountered by the actuator, such as but not limited to the loss of one of the moving elements, the disconnection of one of the moving elements from the kinematic transmission chain, the loss of several elements of the blocking mechanical member (2) or the breakage of a part of the kinematic transmission chain (20).

[0117] like Figure 1As shown, the electromechanical system (3) can integrate various internal sensors to generate signals used by the control circuit (11). These sensors can be of various types, including vibration sensors (13), temperature sensors (14), pressure sensors, or analog or digital probes that measure the magnetic angle of the rotor. The present invention can also use a combination of several signals to generate integrity diagnostics for the active air flow conditioning system.

[0118] Figure 2 An alternative embodiment of the active air flow control system (1) according to the present invention is shown. Figure 1 The difference shown in is that the magnetic motor (11) is controlled by a sensored BLDC system, i.e. the magnetic motor (11) is equipped with a direct or FOC closed-loop control that directly measures its rotor position. To this end, one or more sensors (15) are provided to inform the microcontroller of the precise position of the rotor of the permanent magnet motor (11), which itself generates the desired displacement and torque.

[0119] Figure 3 A variation of the active air flow control system (1) according to the invention is shown. Figure 2 The difference shown in is that the magnetic motor (11) is a sensorless BLDC, that is, the rotor position information of the magnetic motor (11) no longer comes from the position sensor (15), but is estimated based on electrical parameters measured at the terminals of the phase supply line of the magnetic motor (11) by, for example, a measuring device (17) for phase current and / or phase voltage, which measuring device (17) can be equipped with one or more voltage sensors and / or current sensors.

[0120] Of course, these are not the only ways to implement the electromechanical system, and those skilled in the art may readily envision other alternatives for the design and control of the magnetic motor and for the content of the electromechanical system's internal sensors for detecting faults.

[0121] Various types of faults are shown in several examples, and various means for detecting these faults are also proposed, as well as several solutions for extracting singularities from the information provided by the detection means. Finally, several algorithmic implementations for converting detected singularities into tangible error signals are provided. The described embodiments are only for illustrative purposes and in no way limit the scope of the invention, since a person skilled in the art can use some or all of these embodiments and replace the missing parts with solutions known in the art.

[0122] Detailed description of variants of the fault detection algorithm

[0123] An example of an algorithm (200) describing a fault diagnosis of an active air flow control system (1) is given by Figure 4The flowchart given in FIG. 1 shows that the blocking mechanical member of the active air flow regulation system is a set of gates. After successfully completing a single or cyclic learning phase (100), as shown in FIG. Figure 5 As shown, the algorithm may be triggered upon receipt of a request from the DCU (8), or may be automatically initiated when the DCU (8) triggers the actuator to move.

[0124] An optional step (120) upstream of the algorithm consists in obtaining authorization from the DCU (8) to perform a diagnosis of the system state in the case of an on-demand diagnosis, in the event of failure to obtain authorization, the diagnosis will be permanently activated during any movement performed by the mechatronic system.

[0125] The first step (210) consists in starting a movement phase of the blocking mechanical member (2) generated by the electromechanical system (3). During this movement phase, a step (211) consists in acquiring a series of data s by the control circuit (11). i The storage of data may involve partial or complete movement of the blocking member and in any case ends at step (212), triggered by the end of said movement. The microcontroller then uses a dedicated algorithm to analyze the stored data sequence s i To detect singularities (213), this can be similar to a machine learning or deep learning model. Then there is step (214), which is used to i The information obtained from the analysis is classified, resulting in four possible results:

[0126] a. No singularity is detected, so the system is classified as "nominal". This is followed by step (220) in which the electromechanical system (3) generates an operating code for the DCU (8), which code is associated in step (221) with the transmission to the DCU (8) of the confidence level of the previously performed classification (214).

[0127] b. A singular point is detected corresponding to a degradation equivalent to the failure of a single active brake, and the system is therefore classified as a "system with one degraded brake". This is followed by step (230), in which an occasional error code is generated for the DCU (8) by the electromechanical system (3), which code is associated in step (232) with the transmission of the confidence level of the classification previously performed in step (214) and with the transmission of the environmental context in step (233) in which the diagnosis of steps (210) to (214) has been performed.

[0128] c. A singular point corresponding to a degradation equivalent to the failure of several active gates is detected, and the system is therefore classified as a "system with significant degradation". This is followed by step (231) in which an occasional error code for the DCU (8) is generated by the electromechanical system (3), which code is associated in step (232) with the transmission of the confidence level of the classification previously performed in step (214) and with the transmission of the environmental context in step (233) in which the diagnosis of steps (210) to (214) has been performed.

[0129] d. The classification resulting from the analysis of the data series performed in step (214) does not allow us to aggregate information with a sufficient confidence level to be able to distribute it to the DCU

[0130] (8). This is followed by step (240), in which a code corresponding to the "unavailable" state of the diagnostic classification is generated and transmitted by the electromechanical system (3) to the DCU (8), which code is associated in step (241) with the transmission of the environmental context to the DCU (8) in which the diagnostics of steps (210) to (214) have been performed and cannot be assembled.

[0131] If the confidence level is greater than or equal to a preset threshold value (e.g., 99.99%) that depends on the confidence level generated in the previous step, step (250) directs the user to step (260) that is common to all classes providing tangible information. In step (260), an integrated error code or an integrated function code is generated and transmitted by the mechatronic system (3) to the DCU (8), which can be used to feed the OBD strategy. Alternatively, if the confidence level is below the preset threshold value, step (250) is followed by step (270) in which the mechatronic system recommends to the DCU a strategy for integrating sporadic errors, the strategy essentially comprising one or more partial or complete actuator movements.

[0132] In step (300), if step (241) has been performed, the electromechanical system (3) summarizes updates to the analysis of the stored data sequences (ie, the dedicated algorithm or model) to enable classification in the future of data sequences that have not yet allowed the current statistical analysis to be compiled.

[0133] In step (300), if step (270) has been performed, the electromechanical system (3) initiates an updated summary of the analysis of the stored data sequences (i.e., the dedicated algorithm or model) so as to be able to classify the data sequences in the future with the highest confidence level compared to the confidence level just calculated.

[0134] Steps (260) and (300) then lead to step (280) which corresponds to the end of the fault identification algorithm.

[0135] It is noted that the integration strategy may be handled by the algorithm proposed herein for fault identification, or may be an alternative algorithm potentially dedicated to the integration code transmitted to the DCU (8), proposed by the system or applied by the DCU (8). The DCU (8) may be responsible for the priority of the strategy to be adopted to integrate the information (i.e., occasional errors or good working order codes).

[0136] The fault identification algorithm described herein is for illustrative purposes only and should in no way be considered as limiting the scope of the invention. A person skilled in the art can easily conceive of similar alternatives that satisfy the basic technical requirement of informing the DCU (8) of deviations from nominal operation via error codes.

[0137] The sequence (300) describing the back propagation applied to the learning model by example is composed of Figure 5 The flowchart shown in FIG. 3 shows that if it is revealed in step (301) that the previous analysis of the data sequence has diverged, rather than heading towards Figure 4 If convergence on the correct classification of the system among the predetermined options given in , an internal algorithm analysis step (302) is performed by the control circuit (11) (or by the vehicle DCU (8)) in order to propose an update to the model (i.e. the algorithm) used to perform the analysis of the data sequence. In step (303), if an update can be proposed to classify the data sequence while ensuring the same performance level as the classification of the previous samples, a request for a formal update can be sent to the vehicle DCU in step (307), and if the request is approved, the update is stored and activated in the MCU in step (308). However, if no update solution can be identified in steps (302) and (303), then in step (304), the environmental context in which the diagnosis is performed is stored in a non-volatile memory of the control circuit (11) (or vehicle DCU) 8)) to enable the limitations of the current solution to be identified later, which completes the sequence.

[0138] If it is revealed in step (301) that the previous analysis of the data sequence has produced a classification with a confidence level of less than 99.99%, then in step (305) an internal algorithm analysis loop is started by the control circuit (11) (or by the vehicle's DCU (8)) in order to propose an update of the model used to perform said analysis of the data sequence (i.e. the model of the algorithm). In step (306), if an update can be proposed to generate a classification with a confidence level greater than 99.99% while guaranteeing the same confidence level as the classification of the previous sample, a request for a formal update can be sent to the vehicle DCU (8) in step (307) and, if the request is approved, the update is stored and activated in the control circuit (11) in step (308). If no update solution can be proposed in step (306), the sequence is terminated.

[0139] Different types of failures

[0140] Figure 6 An example of an active air flow control system (1) is shown, the blocking mechanism (2) of which comprises a set of shutters. The active air flow control system is positioned on a vehicle grille in front of the radiator and comprises an electromechanical system (3) which drives a shaft (4) passing through an output gear of the electromechanical system (3). The shaft (4) controls the positioning of right shutters (23 to 25) and left shutters (26 to 28) via a transmission (19, 29).

[0141] Figure 7 An alternative example of an active air flow regulation system (1) is shown, the blocking mechanical member (2) of which comprises a frame (31) which can be closed by a curtain (32) which can be wound around a shaft (4) passing through an output gear of an electromechanical system (3).

[0142] Figures 8 to 16 Different types of faults associated with the active air flow control system (1) are shown in a non-limiting manner, wherein Figures 8 to 12 It shows that Figure 6 Failure of the mechanical blocking member (2) equipped with a gate as shown in FIG. Figures 13 to 16 It shows that Figure 7 Failure of a mechanical blocking member (2) equipped with a curtain as shown in . The figures show a system with an active grid whose shutter rotation axis is positioned horizontally, but the invention also extends to any other orientation of the shutter rotation axis.

[0143] Figure 8 The case is shown where one of the center gates is missing due to impact or breakage of its pivot pin. It then gives way to a permanent opening (34) through which air can flow, regardless of the orientation commanded by the actuator. However, the missing gate (24) will result in changes in the forces exerted on the electromechanical system (3) during changes in orientation, and the signatures of these forces can be detected by processing the electrical signals measured directly at the power and control ports of the electromechanical system (3). Figure 8 In the described examples, the central gate is missing, but the present invention is not limited to this case, and different signatures can be observed when end gates are missing or even when several gates are missing, the signature also varying depending on the combination of missing gates.

[0144] Figures 9 to 11Another fault situation on an active flap air flow control system is shown, where all flaps (26, 27, 28) are present, but one (or more) of these flaps is no longer driven by the motion transmission kinematics (36). This could be due, for example, to a break in the connecting rod (37), such as Fig. 9 As shown, or due to the breakage of the gate drive bracket (38), as shown Fig.10 As shown, or due to the breakage of the transmission shaft (39), as Fig.11 Visible in.

[0145] This fault results in a change in the forces exerted on the electromechanical system (3) during the orientation change. The signatures of these forces can be detected by processing the electrical signals measured directly at the power and control ports of the electromechanical system (3).

[0146] Fig.12 Another fault situation is shown in which the gate (24) is no longer driven and interferes with the opening or closing of the adjacent gates (23, 25). This behavior may occur when the drive shaft of a gate breaks, accompanied by a deviation in the stroke of the other gates. This type of breakage may result in a one-time overtorque required to disengage the contacting gates, or a severe overtorque due to the system being completely locked in a position outside the functional stop. It should be noted that the jamming of the drive shaft of one or all dampers can also lead to a one-time or overall increase in the torque required for the opening or closing stroke.

[0147] Figures 13 to 15 Another fault situation of the active curtain air flow control system is shown, in which a component in the motion transmission kinematics has failed. This can be caused by, for example Fig.13 The drive shaft (4) shown is broken, Fig.14 The breakage of the drive cable (33) shown, or Fig.15 The tear (40) of the curtain (32) shown is caused.

[0148] This results in changes in the forces exerted on the electromechanical system (3) during the orientation change. The signatures of these forces are detected by processing the electrical signals measured directly at the power and control ports of the electromechanical system (3).

[0149] Fig.16 Another fault situation is shown, in which the frame (31) of the curtain (32) is deformed or even broken. This type of breakage (35) can lead to a one-time overtorque required to move the curtain, or to a severe overtorque due to a complete blockage of the system. It should be noted that the jamming of the drive shaft of one or all dampers can also lead to a one-time or total increase in the torque required for the opening or closing stroke of the electromechanical system (3).

[0150] Detailed description of the first singular point measurement variant

[0151] In one embodiment, detection of operating singularities is based on sampling and analysis of load angles.

[0152] The load angle corresponds to the magnetic angular displacement of the rotor relative to the angular position that cancels the magnetic torque between the rotor and the stator. A common way to represent this is to use a Fresnel diagram, as shown in Figure 17 by sub-diagrams a, b, c and d, in which the magnetic states of the rotor and stator are represented by vectors. In this diagram, a three-phase motor is represented, and the vectors U, V, W are the magnetic states corresponding to each motor phase supplied with the same voltage.

[0153] In this representation, the load angle (52) corresponds to the angle between the stator vector (50) and the rotor vector (51) of the magnetic field. The resulting torque at the rotor stage produced by the power supply at the stator stage changes from zero torque when this angle is equal to 0° (as Fig.17a The maximum value when the vector is at 90° is shown in Fig.17d The torque represented by the torque vector (53) is proportional to the sine of the load angle (52) and the supply current. Fig.17b and Fig.17c Intermediate cases for load angles between 0° and 90° are shown in order to visualize the evolution of the torque vector between these values.

[0154] Assuming there is no load on the rotor, the load angle (52) is 0°, and the stator vector and rotor vector are collinear. The actual angular position of the rotor is the same as the commanded position.

[0155] When a force such as a braking torque, a load or a driving torque is applied to the rotor, the load angle (52) increases and is no longer equal to 0°. When the load angle (52) exceeds 90°, the applied torque decreases and may cause a loss of synchronization between the rotor magnetic field and the stator magnetic field, known as rotor stall.

[0156] Therefore, measurement of the load angle (52) requires measurement of the rotor angular position.

[0157] In its broadest sense, the invention is compatible with all types of synchronous motor control, but the solution is more specifically dedicated to stepper motor control. Various solutions are available for measuring the rotor position, and the use of dedicated sensors is advantageous, such as analog magneto-sensitive probes, which however have the disadvantage of limited resolution, which shows its limitations when measuring very small changes in the load angle (52). In actuators with high mechanical reduction, such as those used in the present invention, the reduction chain reduces the perceptible impact of the load changes experienced by the member to be driven on the drive motor by a factor equal to the reduction. Moreover, due to this reduction, the rotor speed is greater than the speed of the driven member by a factor equal to the reduction. Therefore, calculating the rotor angle at each step or each microstep requires very large and therefore expensive computing resources, generally incompatible with the target price of these applications. Therefore, instead of measuring the load angle by comparing the stator angle with the absolute rotor angle reconstructed from two signals from the analog orthogonal probe, it is proposed to significantly reduce the number of load angle measurements during a complete rotor rotation to improve the angular resolution.

[0158] A first solution is to continue to use the analog probe to measure the field of the sensor magnet, but without reconstructing the absolute rotor angle, which is a source of inaccuracy and consumes computing resources. We then want to measure only the zero crossings of the probe, which are the most accurate because they are where the field changes the most. In order to correctly identify this zero crossing, it is necessary to measure and store the maximum and minimum values ​​of the probe for each magnetic cycle, and then reconstruct this zero crossing as the average between the maximum and minimum values ​​of the previous step. This ensures insensitivity to changes in the magnetization amplitude, which are slow irreversible changes due to aging, or reversible shorter-term changes due to, for example, temperature changes. Measuring the load angle therefore involves:

[0159] Trigger step or micro-step counter n s ;

[0160] When the analog probe's zero crossing point Z is measured c Stop the counter when , preferably using the formula V M and V m are the maximum and minimum values ​​of the number of revolutions n-1 respectively;

[0161] Storing the maximum and minimum values ​​measured by the analog probe for the number of revolutions n;

[0162] Use formula Calculate the load angle LA , where N s is the number of steps or microsteps in a mechanical cycle.

[0163] The second solution is to replace the analog probe with a digital probe. This solution slightly reduces the accuracy obtained, since it does not allow the transition threshold to be adjusted for each magnetic cycle according to the measurements made during the previous cycle, but offers both financial savings and a reduction in the required computing resources. By integrating a digital probe, a microcontroller without analog inputs can be used and it is no longer necessary to reconstruct the rotor angle by calculation. The method for measuring the load angle is also similar to the previous solution and consists in:

[0164] Triggering the step counter or micro-step counter when the zero crossing of the digital probe is measured;

[0165] Stop the counter when the digital probe crosses zero.

[0166] Use formula Calculate the load angle L A .

[0167] For these two solutions, the following formula L A,1t To average the load angle measurements over the electrical cycle:

[0168] L A,1t =n s1 +n s2 ,

[0169] n s1 and n s2 are the number of steps counted when the counter is triggered at a predetermined rotor angle

[0170] For higher accuracy, the motion can be analyzed over a large number of electrical cycles or even over one or more rotations of the gearbox output shaft. In this case, if more weight is to be given to the most recently measured samples, a sliding average of the load angle over N samples can be used such as a weighted sliding average:

[0171]

[0172] L A,1t (x) is the average load angle measurement for the xth electrical revolution, and L A,1t (k) is the load angle measurement for the xth electrical revolution. It should be noted that good results are obtained with a number of samples N between 50 and 100. Of course, other types of sliding averages may be considered, depending on the required accuracy and the available memory and computing resources.

[0173] Detailed description of the second singularity measurement variant

[0174] In one embodiment, the detection of operating singularities is based on the sampling and analysis of electrical signals measured at the terminals of an RC filter connected in series with the power supply of the electromechanical system (3), also known as an actuator, which comprises a multiphase motor coupled to a so-called self-commutated control, known to those skilled in the art as "BLDC control" and carried out by an electronic module integrated into the electromechanical system (3).

[0175] Fig.18 A simplified circuit diagram of a three-phase driver according to the invention is shown, which uses six transistors Q1 to Q6, namely Q1 and Q4 (respectively Q2 and Q5, Q3 and Q6) to drive current through phase C (respectively B and A). Resistor (60) is a sampling resistor for measuring the sum of the currents flowing through the phases of the magnetic motor (10) via an RC filter (63) consisting of a resistor (61) and a capacitor (62). The output of the RC filter (63) is acquired as a voltage by an analog / digital converter and continuously transformed and preprocessed into information, for example by a microcontroller.

[0176] A brief description of the measurement alternatives to singularities

[0177] Another parameter that may be of interest to measure for fault detection is the number of steps taken by a motor driven in stepping mode to move from one stop position to another. This type of magnitude is particularly useful for detecting jamming faults, as the effective stroke can be greatly reduced. In a detailed example, this parameter can also be used to identify a missing gate.

[0178] The measurement of the singularity is not limited to functional quantities in the electric motor, such as current measurement or load angle measurement as described above, but can easily come from dedicated sensors inside the actuator. An example is a pressure sensor or a temperature sensor, the variation of which will be proportional to the incoming air flow. It is therefore entirely conceivable to develop different scenarios to make the measurements of these sensors sensitive to the above mentioned faults.

[0179] Description of an example of the singularity identification algorithm

[0180] For example, with Fig.18One means of singularity detection compatible with the current measurements shown in the description of is to identify values ​​in the stored data sequence that exceed the expected value by more than a number of standard deviations. This is described here using a formalism developed for current measurement, but can also be used for other types of measurement data. Singularity detection is achieved by digitally processing the evolution of a physical value consisting of the sum of the currents I of the N phases of the multiphase motor measured in the resistor (60). During nominal operation in synchronous mode, the current has a stable value that is regular and without symptomatic behavior of the rotor of the geared motor, but when the geared motor is affected by an abnormal or faulty active air flow control system (1), this value evolves towards a different and more irregular value.

[0181] For example, a microcontroller's analog / digital converter is used to sample the processing of changes in the current I measured in the resistor (60) and calculate the standard deviation σ of the amplitude values ​​of the current measured across the resistor (60) over N samples over a sliding or fixed time window that is less than or equal to the duration of the movement associated with the acquisition of the information.

[0182] The characterization of the operating mode and the detection of faults are performed by analyzing the variation of this standard deviation σ, possibly supplemented by analyzing the evolution of the median value over N samples of the current measured at the terminals of R1:

[0183] - A stable standard deviation σ below the threshold measured during nominal operation corresponds to trouble-free operation

[0184] - An increased standard deviation exceeding a threshold value and / or a significant increase in the median current measured at the terminals of R1 indicates a sticking of one or more gates and / or their drive mechanisms and, therefore, a fluid control drift

[0185] An increased standard deviation exceeding a threshold value and / or a significant reduction in the median current measured at the terminals of R1 corresponds to a loss of one or more gates, being stuck in an open position, or a partial destruction of one or more gates and therefore to a drift in the fluid control.

[0186] - As an example, the value N 2* σ 2 is compared with the total current value I of the phase. This digital processing makes it possible to distinguish between normal operating areas and abnormal operating areas and to obtain the value N 2* σ 2 A standard deviation threshold E is set above which the microcontroller or ASIC determines that the geared motor is driving a faulty gate mechanism. A combination of several other types of algorithmic statistical analysis models may also be used.

[0187] Description of variants of the singularity identification algorithm

[0188] In one embodiment, singularities are identified using a learning algorithm that can be trained to recognize the type of error encountered, such as a missing gate or a broken gate blocking the system, but can also identify the location of the fault, such as a missing gate at one end or in the middle.

[0189] If the component to be driven includes a set of gates, the fault detection algorithm may benefit from modifying the component to be driven to improve detection. Therefore, it is envisaged to provide each gate with a specific signature that can be reliably identified by the algorithm. As in patent application US9810138B2, this specific signature can be a binary code, in which a point resistance element is added to each gate, which can be detected in the form of a point over-torque generated by the actuator, and the point resistance element is wisely placed so that each causes over-torque at a different moment of the opening or closing stroke. The absence of a gate is then indicated by the absence of a friction peak during the complete movement of the component to be driven. It should be noted that such a signature can be measured by a very simple algorithm, which, for example, detects that the current peak consumed by the actuator exceeds a specific threshold at a predetermined position to approve the presence of each gate.

[0190] If the actuator is a set of flaps or curtains, the signature can be more refined by providing each actuator flap with a specific surface area, a specific mass, or a specific profile, leveraging the power of deep learning. This can lead to the measurement of a very specific signature associated, for example, with the inertia of the system, its pneumatic pressure, or even by obtaining vibrations specific to each flap.

[0191] Description of other variants of the singularity identification algorithm

[0192] It is noted that there are many possible variants of the singularity identification algorithm that are compatible with the present invention, each with its advantages and disadvantages. For example, some are more robust but require a specific actuation sequence of the device to be driven in order to measure the fault; others must be done at rest, while others must be in motion; and still others require the addition of external data such as vehicle speed, wind speed or temperature, so that the selected solution or the selected set of solutions is defined by the specification. Some examples of identification algorithms are given below.

[0193] According to a variant of the invention, the singularity identification algorithm is based on the measurement of the variation of the hydrodynamic torque. When a shutter is missing or a curtain is torn, the pressure exerted on the entire blackout system decreases, resulting in a decrease in the torque required by the actuator to perform an opening or closing movement. The hydrodynamic torque T d It can be expressed as a function of the actuator stroke x according to the following formula:

[0194]

[0195] where c1 is a coefficient related to the geometry of the shading device and the inlet air flow rate, A max is the maximum opening area, and A(x) is the opening area for stroke x, which is expressed differently for a healthy or broken actuator.

[0196] The algorithm can then use the measurement of torque, current or load angle directly as a sequence of sampled digital data in order to identify singularities, or, to increase its accuracy, acquire this data along the actuator stroke in order to identify singularities W(x0,x1) in the work provided by the actuator during the movement between e x0 and x1 according to the following formula:

[0197]

[0198] The measured work can then simply be compared with a reference curve corresponding to a healthy actuator. It should be noted that this method consumes very little computing resources but is highly dependent on wind speed and vehicle speed. However, if the available computing power allows the implementation of a smarter algorithm, several specific actuator movements can be performed to estimate the relative speed of the air impinging on the shielding device.

[0199] According to another variant of the invention, when the shielding system is a set of shutters, the singularity identification algorithm can be based on a measure of the stiffness of the system.

[0200] According to another variant of the invention, the singularity identification algorithm may be based on the measurement of the inertia of the system. If the AGS actuator is controlled so that the angular velocity of the actuator rotor increases linearly between two moments, the rotor acceleration is constant. By performing an opening cycle with the vehicle stationary, a fault, such as a missing gate, can be detected by measuring the change in inertia during this cycle compared to a reference value stored in the system. Alternatively, this change in inertia can be measured from torque, current or load angle readings, but these examples are not limitations of the invention.

[0201] Detailed description of the learning algorithm variants

[0202] In one embodiment, the algorithm describing the fault detection diagnostic requires an initial supervised learning phase that is fed a set number of cycles with the system only in nominal (or functional) mode. An example of such a supervised learning algorithm is Fig.19 The flow chart is shown in FIG.

[0203] We denote the classification system using the data sequence of interest s iThe number of learning cycles X required to provide results with optimal confidence over the entire duration of the sampling acquisition of X. Confidence is a variable that increases as a function of X, grows rapidly as a function of X, and then changes only slightly above a certain value of X. It is therefore possible to obtain confidence close to 100%, with 100% indicating complete certainty and 0% indicating complete uncertainty, with a finite number of learning cycles, such as 10 or 20 depending on the sensitivity of the algorithm, and a desired high confidence level of over 99%.

[0204] Each time the actuator is set into motion by the control circuit (11), the learning algorithm (100) is triggered, which is referred to as step (110), followed by step (111) to check that the initial learning phase has been completed, which can be done by reading the memory registers. If the learning phase has been successfully completed, the learning algorithm terminates and triggers the fault detection algorithm (200).

[0205] If, on the other hand, the learning phase is considered not yet complete, a "non-operating" state of the fault detection system is generated for the DCU (8) by the electromechanical system (3) in step (112). The movement is then initiated by the electromechanical system (3). Step (113) of the learning algorithm consists in sampling the data of interest by the control circuit (11) and in storing these data in an internal memory of the electromechanical system (3). Step (113) is completed when the end of the actuator movement is detected. Then, in step (114), the learning algorithm triggers a statistical analysis of the data stored in step (113) for this learning cycle. The next step (115) consists in checking, by the control circuit (11), the correctness of the movement performed by the actuator and the correct environmental conditions (i.e. the correct context), in order to decide whether to keep the data from the current cycle as reference data for the learning algorithm. If so, these data are stored as reference data in a memory register in a final step (116). The reference data store preferably includes contextual data from the stored cycle, such as information about temperature, vehicle speed, air pressure at the gate, etc., which may come from external sensors, and then associated information potentially generated by the electromechanical system (3) to the DCU (8). If the check concludes that the data is insufficient, the data is not stored, so the cycle is not considered a reference cycle, and the information can be generated by the electromechanical system (3) to inform the DCU (8).

[0206] Implementation scheme for sharing data processing between ECU and DCU(8)

[0207] In the above description, it is generally assumed that all algorithmic processing is performed only via the control circuit ( 11 ), which then transmits information characterizing the nature of the fault detected in the air flow chain (with or without a mechanical fault).

[0208] However, it is also possible to divide the algorithm processing between the control circuit (11) and the vehicle's DCU (8), which usually has a much higher computing power and can optionally also take into account other vehicle data in order to achieve a more complete integration of vehicle diagnostics. However, this embodiment must take into account the fact that the transmission bus between a piece of equipment and the vehicle's DCU has a limited bandwidth and does not allow the raw sensor data to be transmitted directly to the vehicle's DCU. A "distributed" embodiment of the invention then consists in pre-processing the raw data on the control circuit (11) at a high frequency to calculate pre-processed data sets, which are then transmitted from the control circuit (11) to the vehicle's DCU (8) at a frequency compatible with the bandwidth of the communication line (7) (for example, using the LIN protocol), the vehicle's DCU using these data sets transmitted at a low frequency to perform post-processing at a second frequency lower than the first frequency, so as to enable the DCU to complete the analysis and identify sporadic errors, which then enable the DCU (8) to characterize pneumatic and / or mechanical faults of the air flow control system.

Claims

1. A method for diagnosing the operation of an active air flow control system (1) of a vehicle, the active air flow control system comprising: - blocking mechanism (2); - an electromechanical system (3) provided with a housing housing a magnetic motor (10) driven by a control circuit (11) and coupled to movement conversion means for moving the blocking mechanical member (2); - an electronic communication line (7) for exchanging information with the DCU (8) of the vehicle or receiving commands from the DCU (8) of the vehicle, Characterized in that the method comprises a series of steps, wherein the series of steps comprises: - storing in a memory a sequence S of values ​​of data acquired by the control circuit (11) inside the electromechanical system (3) during an operating cycle of the active air flow regulation system (1); - applying an algorithmic statistical analysis model to said sequence of values ​​S of data stored in said memory in order to determine singularities; - Identifying sporadic errors associated with said singularity in order to provide information about the state of said active air flow control system (1).

2. Method for diagnosing the operation of an active air flow control system (1) according to the preceding claim, characterized in that The step of identifying the sporadic error is performed by the control circuit (11), which transmits information representing the identified sporadic error to the DCU.

3. Method for diagnosing the operation of an active air flow control system (1) according to claim 1, characterized in that The sequence of digital data stored in the memory is the subject of the following operations: - pre-processing by the control circuit (11) at a first processing frequency; and - post-processing by the DCU / ECU (8) of the vehicle at a second processing frequency lower than the first processing frequency to complete the analysis and identify sporadic errors; The data set preprocessed by the control circuit (11) is transmitted to the DCU / ECU (8) via the command line (7).

4. Method for diagnosing the operation of an active air flow control system (1) according to any one of the preceding claims, characterized in that The digital data sequence S stored in the memory comprises: a. Data of the first type; and b. At least a second type of data different from the first type.

5. Method for diagnosing the operation of an active air flow control system (1) according to any one of the preceding claims, characterized in that The method comprises the following steps as additional steps: a. integrating the sporadic errors according to instructions transmitted by the DCU (8) of the vehicle, thereby generating an integrated diagnostic status; b. Transmitting the integrated diagnostic status to the vehicle's DCU (8).

6. Method for diagnosing the operation of an active air flow control system (1) according to claim 1 or 5, characterized in that One of the data is a measurement of the motor phase current.

7. Method for diagnosing the operation of an active air flow control system (1) according to claim 1 or 5, characterized in that One of the data is a measurement of the load angle of the rotor of the magnetic motor (10).

8. Method for diagnosing the operation of an active air flow control system (1) according to claim 1, characterized in that The statistical analysis algorithm is provided with a learning sequence.

9. Method for diagnosing the operation of an active air flow control system (1) according to claim 1, characterized in that The statistical analysis algorithm is provided with a cluster selection method.

10. Method for diagnosing the operation of an active air flow control system (1) according to claim 1, characterized in that The blocking mechanism (2) has no diagnostic-specific components external to the electromechanical system (3).

11. Method for diagnosing the operation of an active air flow control system (1) according to claim 1 or 5, characterized in that One of the data relates to the operation of the magnetic motor (10) of the electromechanical system (3).

12. Method for diagnosing the operation of an active air flow control system (1) according to claim 1, characterized in that The algorithm model for statistical analysis to determine singularity is set up with a sequence S of stored sampled digital data and a reference data map S r Comparison between.

13. Method for diagnosing the operation of an active air flow control system (1) according to claim 1, characterized in that The algorithmic model for statistical analysis to determine singular points is provided with the detection of at least one singular point in the sequence S of stored sampled digital data.

14. Method for diagnosing the operation of an active air flow control system 1 according to claim 1, characterized in that The algorithmic model for statistical analysis to determine singularities consists in subjecting the sequence S of stored sampled digital data to a model obtained by training a neural network according to training data corresponding to the expected operation of the blocking mechanical member (2).

15. Method for diagnosing the operation of an active air flow control system (1) according to one of the preceding claims, characterized in that A reference sequence S of sampled digital data is stored for each electromechanical system (3) at the end of an assembly line of the system (1) or after the system (1) is installed on the vehicle. r .

16. Method for diagnosing the operation of an active air flow control system (1) according to the preceding claim, characterized in that The algorithm model for statistical analysis can compare the sampled digital data sequence with the reference sequence S of sampled digital data. r The comparison is performed in order to detect behavioral drifts that could cause degradation of the electromechanical system or to verify normal wear behavior of the system.

17. Method for diagnosing the operation of an active air flow control system 1 according to claim 1, characterized in that The stored sequence S of sampled digital data is derived from digital data that depends on the mechanical load in the active mesh system.

18. Method for diagnosing the operation of an active air flow control system 1 according to the preceding claim, characterized in that Identifying singular points in a sequence S of stored sampled digital data is accomplished by knowing: · Mass or inertia to be moved in the active grid system; and / or mechanical friction in the system; and / or pneumatic loads imposed on the system; and / or system temperature.

19. An electromechanical active air flow control system (1) for a vehicle, comprising: - blocking mechanism (2); - an electromechanical system (3) provided with a housing housing a magnetic motor (10) driven by a control circuit (11) and coupled to movement conversion means for moving the blocking mechanical member (2); - an electronic communication line (7) for exchanging information with the vehicle's DCU (8) or receiving commands from the vehicle's DCU (8), characterized in that the method comprises a microcontroller or microprocessor executing a computer program stored in its read-only memory, controlling a series of steps corresponding to the preceding claims.

Images