EP1978490A1

Movatterモバイル変換

Info

Publication number: EP1978490A1
Application number: EP07425194A
Authority: EP
Inventors: Piero Mortara
Original assignee: Magneti Marelli Sistemi Elettronici SpA
Current assignee: Marelli Europe SpA
Priority date: 2007-04-02
Filing date: 2007-04-02
Publication date: 2008-10-08

Abstract

A description is given of a system and a method for the recognition of the operating state of a vehicle engine, based on the detection of an acoustic noise signal, and possibly a vibration signal, propagated in a compartment of the vehicle, and on the comparison of a descriptor vector of the detected signals, comprising at least data indicative of the frequency and energy components of the signals, with a plurality of reference descriptor vectors indicative of known operating conditions, acquired in a learning phase, in such a way as to discriminate and recognize a first class of signals indicative of an operating condition and a second class of signals indicative of an inoperative condition of the vehicle engine.

Description

The present invention relates in general to the monitoring of the operating conditions of a vehicle, and more specifically it relates to a system and a method for the automatic recognition of the started and stopped conditions of a motor vehicle engine.
Whenever a vehicle has to be fitted, after its manufacture, with an electronic device which requires the detection of the transition to operation (engine starting or "Key-On") and vice versa (engine stopping or "Key-Off"), the problem arises of integrating the device in the vehicle for the interception of the vehicle operation enabling signal, and consequently the recognition of the operating state of the engine, in a relatively non-invasive way.
There are known systems for recognizing the operating state of a vehicle engine which require an electrical connection to the on-board electrical distribution system, for example to the starting switch device of the vehicle, such as the ignition key unit or other similar device used to start the vehicle.
This connection usually causes problems, since it requires the identification of the location of the starter switch, which can differ substantially with the type of vehicle (motor car, industrial vehicle, articulated lorry, motorcycle, agricultural or earth-moving vehicle, ...) and with the technology used for the starting function (determined, for example, by the age of the vehicle). For example, in more recent vehicles the mechanical starting key is sometimes replaced by an electronic device of the transponder type (such as a smart key or a smart card), and in this case the interception of the enabling signal for the operation of the vehicle becomes even more problematic.
The search for points of connection to the electrical distribution system of the vehicle, and the installation of the wiring, requiring the dismounting and refitting of some of the vehicle's equipment (such as the dashboard, the control panel, the steering unit, etc.), are time-intensive operations with consequential effects on working costs, and make it necessary to keep the vehicle in the workshop. Furthermore, the intrusiveness of the operation is not only unappealing to the client but can also have repercussions on the reliability and guarantee of the vehicle, if the operation is not carried out in suitable conditions.
The object of the present invention is to provide a satisfactory solution to the problems described above, while avoiding the drawbacks of the known art; in other words, to provide a method of automatic recognition of the operating condition of a vehicle engine for an electronic device to be installed or positioned on board subsequently, which does not require a physical connection with the on-board systems of the vehicle.
According to the present invention, this object is achieved by means of a system having the characteristics claimed inClaim 1, and a method having the characteristics claimed inClaim 20.
Specific embodiments are described in the dependent claims.
Briefly, the present invention bases the recognition of the operating conditions of a vehicle engine on the analysis of the acoustic noise signals (in other words, signals which are different from voice signals) and if necessary the vibration signals present in at least one compartment of the vehicle such as the passenger compartment or the engine compartment, detected by means of sensors (for example, low-cost sensors such as microphones and accelerometers) and processed by an electronic unit (such as a microprocessor, a digital signal processor (DSP), or an FPGA wired computational logic or other specific ASIC logic device) adapted to carry out, in a cooperative mode, a spectrographic analysis (in other words an analysis of the frequency components of the signal), a dynamic analysis (in other words an analysis of the intensity and therefore the energy of the signal) and possibly a phase analysis (phase shifts between copies of the signal detected at different locations on board the vehicle).
The validation of the recognized conditions can also be additionally associated with (or supported by) the detection of the movement of the vehicle by means of additional accelerometers and/or satellite location signals, for example GPS.
This cooperative noise signal analysis also makes it possible to extrapolate information on the current rotation speed of the engine, thus making this information available for further processing.
The system can be constructed either as an independent device capable of providing the recognition of the On/Off condition (vehicle operational/vehicle non-operational) in the form of output information which can be used by other devices, or as a system integrated in an electronic device having any function which requires the recognition of the aforesaid condition.
The engine speed information, appropriately correlated with position and time data, as supplied for example by a GPS receiver, makes it possible to derive operating parameters of the vehicle for the execution of advanced mobility control functions.
For example, it is possible to use a subscriber's on-board communication device to detect access, determine the period of presence and the conditions of use of a vehicle in a restricted traffic area (for example, a road charging area), in order to determine by a weighted method any pricing to be applied, which can then be calculated in accordance with the modes of access and use of the vehicle.
For example, if the reduction of polluting emissions is to be rewarded, the level of pricing for access to a limited traffic zone could be determined according to the time for which the vehicle is present in this zone with the engine running, and according to the conditions of use, which become more or less expensive according to the engine speed and the driving speed.
In this case, the pricing can be applied on the basis of an identifier of the user and an identifier of the vehicle, communicated to an access control system by means of an external programmed module, or read directly in wireless mode from a transponder device, a smart card, or the like.
Advantageously, the described method requires no physical connection to the vehicle system (to the electrical distribution system, the engine speed sensor, electronic control units, the on-board data network, etc.), and no calibration or adjustment operation, since it uses a self-adapting algorithm which can automatically optimize itself after an initial learning phase. By obtaining the aforesaid information without the need for a physical connection to the vehicle system, it is possible to avoid complicated and intrusive installation operations and their implications in terms of time, cost and potential effects on the reliability and guarantee conditions of the vehicle.
Further advantages and characteristics of the invention will be made clear by the following detailed description, which refers to the attached drawings provided purely by way of example and without restrictive intent, in which:

Figure 1 is a block diagram of the system proposed by the invention;
Figure 2 is a logic diagram indicating the functions of the system proposed by the invention;
Figure 3 shows a set of graphs representing the analysis functions carried out by the system proposed by the invention;
Figure 4 is a state diagram of a self-leaming procedure of the system proposed by the invention;
Figure 5 is a state diagram of a guided learning procedure of the system proposed by the invention;
Figure 6 is a state diagram of a recognition procedure of the system proposed by the invention; and
Figure 7 is a schematic representation of a method of monitoring the conditions of use of a vehicle, based on the system for recognition of the engine operating state proposed by the invention.

Figure 1 shows an exemplary block diagram of the system for recognizing the operating state of a vehicle engine, for example, but not exclusively, an internal combustion engine, according to the invention. This system can be provided in a portable electronic device, or a device to be permanently installed or positioned on board a vehicle, since the system can be efficiently integrated in its own circuit board having small overall dimensions (if it forms a stand-alone device, for example) or integrated in a pre-existing circuit board of a more complex device (for example, a processing and communication box for monitoring the driving behaviour).
Thenumber 10 indicates a processing unit, for example an ordinary microprocessor-based processing unit, a digital signal processor (DSP), an FPGA (Field Programmable Gate Array) wired computational logic or other specific ASIC (Application Specific Integrated Circuit) logic device which can execute a software or firmware program.
This unit is adapted to receive at a first input electrical signals arriving from at least oneacoustic sensor 12 and preferably from at least a pair of these sensors, for example microphones operating at audio frequencies, which are adapted to acquire acoustic or noise signals propagating in a reference compartment of the vehicle, such as the passenger compartment, and from at least onevibration sensor 14 of the vehicle, such as an accelerometer or similar transducer mechanically coupled to a rigid part of the vehicle, which is adapted to detect vibrations of the vehicle, particularly at subsonic frequencies, thus effectively extending the analysis spectrum.
The electrical signals are taken to theunit 10 via an A/D converter module 16.
The unit is also adapted to receive, at a second input, signals sent by an on-boardgeographical positioning system 18, for example a GPS satellite positioning system.
Memory means 20 are associated to theunit 10, adapted to store a plurality of reference models of acoustic noise signals and vibration signals classified according to known operating conditions, for example with reference to an operating condition in which the engine is started and an operating condition in which the engine is stopped.
Theprocessing unit 10 is also designed to receive at its input supplementary data for identifying thevehicle 30 and possibly the user of thevehicle 32, using known means of communication.
Theprocessing unit 10 includes a finite state machine which, in a learning phase, is adapted to carry out the generation of a plurality of reference models of acoustic noise signals and vibration signals (if required, for the extension of the analysis to subsonic frequencies), and, in an execution phase, the classification of the acoustic noise signals and the vibration signals received at the input according to the said plurality of reference models. It is designed to send an output electrical signal indicating an operating state of the vehicle, for example a binary signal to signal the operating condition in which the engine is started or in which the engine is stopped.
A detail of the logical architecture of the processing unit is described with reference toFigures 2 and3.
Theprocessing unit 10 comprises afirst processing module 40 adapted to carry out a spectrographic analysis of the acoustic noise signal received (and of the vibration signal, if specified) in order to generate frequency parameters indicative of the said signal, asecond processing module 42 adapted to carry out a dynamic analysis of the acoustic noise signal received (and of the vibration signal, if specified) in order to generate dynamic parameters indicative of the said signal, and athird processing module 44 adapted to carry out a phase analysis of the acoustic noise signal received (and of the vibration signal, if specified) in order to generate phase parameters indicative of the said signal.
The frequency, dynamic and phase parameters are combined to form a descriptor vector of the detected signal.
The frequency parameters comprise, for example, a vector for spectral representation of the signal components, as shown inFigure 3, the spectrographic analysis comprising a Fourier analysis of the noise signal carried out through digital filters, for example in order to separate or exclude voice signals or other spurious noise signals (external to the vehicle). In fact, the spectrographic analysis makes it possible to highlight deterministic periodic components of the signal, produced by the engine, and separate them from nondeterministic spurious noises.
Alternatively, the voice signal can be separated or excluded by filtering upstream of the analysis processing modules 40-44, or downstream of them, in the recognition algorithm.
The dynamic parameters comprise, for example, a vector of data indicating the intensity and energy of the signal over time, as shown inFigure 3, the dynamic analysis comprising an analysis of the envelope of the noise signal received.
The phase parameters comprise, for example, a vector representing the phase shift components between copies of the noise signal received from sensors distributed in the compartment, as shown inFigure 3. The phase analysis makes it possible, for example, to separate or exclude voice signals or other spurious noise signals external to the vehicle, which may be similar but are probably not repeated over a long period.
Acomposition processing module 46 downstream of the modules 40-44 is adapted to receive the corresponding analysis signals from the said modules and compose a global model of the detected signal as a function of the frequency, dynamic and phase parameters calculated previously.
In a learning phase of the system, the global model composed by themodule 46 is stored in alibrary 48 of reference models in the memory means 20. In an execution phase of the system, the global model composed by themodule 46 is compared with the reference models in thelibrary 48 in a comparator module 50, which is arranged to recognize the current operating condition and to send a signal indicative of this condition.
The comparison is made by determining the distance between the descriptor vector of the detected signal and the reference descriptor vector, two vectors being considered similar if the distance between them is less than a predetermined threshold value. This threshold value is determined by identifying a predetermined acceptable range of variance for each component of the vector which can be obtained by the aforesaid analyses, centred on an expected reference value.
Figure 2 shows both the learning phase and the execution phase, but it should be understood that the elements of the system shown twice are not actually duplicated in the system, their representation being duplicated solely to aid the understanding of the functions of the system.
The comparator module 50 can also be conveniently provided with input signals indicative of a condition of movement of the vehicle, from either anaccelerometer device 60 or the integratedGPS system 18, respectively.
Timer means are also integrated into theprocessing unit 10 for controlling the wake-up of the unit at predetermined time intervals, for performing a recognition operation. Alternatively, this wake-up could be performed after the unit has received an external wake-up signal.
Figure 4 shows a state diagram of theprocessing unit 10 in an autonomous learning phase.
Theunit 10 is woken up from aninactive state 100 periodically (for example, every 30 seconds), and then switches to anactive state 120.
In this state, it checks whether the acoustic signal level perceived by the sensors is or is not above a predetermined minimum threshold. If the perceived acoustic signal is below this threshold, the unit resets itself to theinactive state 100 and waits for the next wake-up. If there is a signal above the threshold, the unit starts asignal analysis step 140 supported by predetermined reference models, such as known models dependent on the nominal characteristics of the vehicle engine type.
If there is a representative signal, in other words - in the space of the signal descriptor vectors - one lying within a neighbourhood with a predefined radius of the descriptor vector of a given nominal reference model of operating conditions for the type of engine or vehicle, its parameters are stored (step 140), and the library of reference models is updated accordingly. Otherwise, the unit resets itself to theinactive state 100. It always switches to this state after the updating of the model library.
Figure 5 shows a state diagram of theprocessing unit 10 in a guided learning phase.
Theunit 10 is woken up from theinactive state 100 switching to theactive state 120 when the engine is started, in other words when the key unit is switched.
Theunit 10 then starts a step of analysis of thesignal 140 and of storage of the parameters, updating the library of reference models accordingly, and then resets itself to theinactive state 100.
Figure 6 shows a state diagram of theprocessing unit 10 in an execution step of recognition of an operating condition of the vehicle engine.
Theunit 10 is woken up from theinactive state 100 periodically (for example, every 30 seconds), and then switches to theactive state 120.
In this state, it checks whether the acoustic signal level perceived by the sensors is or is not above a predetermined minimum threshold. If the perceived acoustic signal is below this threshold, the unit resets itself to theinactive state 100 and waits for the next wake-up. If there is a signal above the threshold, the unit starts asignal analysis step 140, supported by the library of predetermined reference models, as described above.
If the signal is recognized, the unit sets itself to a permanentactive state 200 and sends a signal indicative of the recognized operating condition. From this state, it performs a newsignal analysis step 140 periodically (for example, every 30 seconds).
When, after asignal analysis step 140, the unit no longer recognizes the acoustic signal perceived by the sensors, it resets itself to theinactive state 100 possibly, sending a signal indicative of the recognized operating condition.
Advantageously, the disclosed system can be integrated into a subscriber's on-board device forming part of a system for managing access to a controlled-access area, such as an urban area with restricted traffic.
Permission to access and freely pass across an area of the aforesaid type is normally granted to vehicles with which a specific transit permit has been associated because they belong to specific categories due to the fact that they are equipped with particular devices, for example devices capable of limiting polluting emissions, or are intended to provide particular services. These permits can generally be allocated to vehicles allowed to travel to a predetermined destination lying within the area, for example vehicles belonging to members of the public residing in the area itself, or having their workplaces there.
To regulate the passage of such vehicles, use is commonly made of special road signs which indicate the limits of the "restricted traffic" areas, and of penalties imposed on offenders by traffic police, which form crude, inefficient and inflexible control systems.
Alternatively, physical barriers have been proposed, controlled by automatic systems to allow the passage of only those vehicles which have been previously authorized to pass through the barriers.
Recently, the use of remote control systems has become established for monitoring the passage of vehicles through points of access to the restricted traffic areas, these systems being adapted to acquire images or codes identifying a vehicle in transit for checking the authorization for access or for pricing the access, for example according to the time for which the vehicle remains in the area.
Figure 7 shows a system for managing access to a restricted traffic area based on the use of on-board devices provided with a system for the automatic recognition of the operating state of the engine of a vehicle according to the invention, for monitoring the conditions of use of the vehicle.
A state diagram of theprocessing unit 10 in an execution phase of recognition of an operating condition of the vehicle engine is shown, as inFigure 6. In the permanentactive state 200, the unit carries out a cyclic activation with a period which is at least 2 orders of magnitude shorter than the wake-up period (for example, 5ms), to start asignal analysis step 240, in which, in a preferred embodiment, the current engine rotation speed is recognized and a corresponding signal rpm is sent, by the filtering or application of processing algorithms with configuration parameters predetermined according to the type of vehicle engine.
In a monitoringrecord compilation step 260, theunit 10 receives in input signals identifying the vehicle and the user from an external programming device (for example, an on-board transponder carrying the said information recorded on a smart card), together with signals indicative of the period of use and the driving speed of the vehicle, using for example a resident GPS system, and records the vehicle use parameters by correlation of the data on the engine rotation speed, period of use, location and speed of the vehicle.
The stored data are transmitted (step 280), for example by means of a wireless connection via GPRS, Wi-Fi or the like, to a service controller of the system for managing access to a restricted traffic area, and this controller applies a price (300) and makes a charge (320) according to the data received. Alternatively, the pricing and charging operations could be carried out locally, for example to carry out local charging to a prepayment card, as shown in the figure.
Thus the pricing and corresponding charging are made strictly dependent on the conditions of use of the vehicle, with particular attention paid to the period for which the vehicle engine is running, and, in a preferred embodiment, to the operating conditions of the engine.
Naturally, the principle of the invention remaining the same, the forms of embodiment and details of construction may be varied widely with respect to those described and illustrated, which have been given purely by way of non-limiting example, without thereby departing from the scope of protection of the present invention as defined by the appended claims.

Claims

A system for recognizing the operating state of a vehicle engine, comprising, in combination,
- acoustic sensor means (12) for detecting an acoustic noise signal propagated in a compartment of the vehicle and for transducing the said signal into a corresponding electrical signal; and
- electronic processing and recognition means (10) of the automatic learning type, arranged for:
receiving the said electrical signal and generating at least one descriptor vector of the detected acoustic signal, comprising data indicative of the frequency components of the signal and data indicative of the energy of the signal;
comparing the said descriptor vector of the detected acoustic signal with a plurality of reference descriptor vectors (48) indicative of known operating conditions, acquired in a learning phase; and
classifying the said detected acoustic signal according to the distance between the descriptor vector of the detected acoustic signal and the reference descriptor vector, so as to discriminate between a first class of signals indicative of an operating condition of the vehicle engine and a second class of signals indicative of an inoperative condition of the vehicle engine.
System according to Claim 1, in which the said descriptor vector of the detected acoustic signal comprises data indicative of a phase shift between at least a pair of signals received by separate elements of the said sensor means, displaced in the compartment of the vehicle.
System according to Claim 1 or 2, in which the said acoustic sensor means (12) comprise at least one microphone device.
System according to Claim 1, 2 or 3, comprising vibration sensor means (14), which are adapted to be coupled mechanically to a rigid part of the vehicle and are adapted to detect a vibration signal propagated in a compartment of the vehicle and transduce the said vibration signal into a corresponding second electrical signal,
in which the said electronic processing and recognition means (10) are also arranged for:
receiving the said second electrical signal and generating at least one descriptor vector of the detected vibration signal, comprising data indicative of the frequency components of the signal and data indicative of the energy of the signal;
comparing the said descriptor vector of the detected vibration signal with a plurality of reference descriptor vectors (48) indicative of known operating conditions, acquired in a learning phase; and
classifying the said detected vibration signal according to the distance between the descriptor vector of the detected vibration signal and the reference descriptor vector, so as to discriminate between a first class of signals indicative of an operating condition of the vehicle engine and a second class of signals indicative of an inoperative condition of the vehicle engine.
System according to Claim 4, in which the said vibration sensor means (14) comprise at least one accelerometer device.
System according to Claim 4 or 5, in which the said descriptor vector of the detected vibration signal comprises data indicative of a phase shift between at least a pair of signals received by separate elements of the said vibration sensor means, displaced in the compartment of the vehicle.
System according to any one of the preceding claims, additionally comprising motion sensor means (18, 60) adapted to detect a condition of movement of the vehicle.
System according to Claim 7, in which the said motion sensor means comprise at least one accelerometer device (60).
System according to Claim 7, in which the said motion sensor means comprise at least one geographical positioning device (18).
System according to any one of the preceding claims, in which the said processing and recognition means (10) are adapted to compare a descriptor vector of the detected signal and a reference descriptor vector according to a predetermined acceptable range of variance for the magnitude of each component of the vector centred on an expected reference value, two vectors being considered similar if the distance between them is less than a predetermined threshold value.
System according to any one of the preceding claims, in which the said processing and recognition means (10) are arranged for estimating the current engine rotation speed by processing the detected signals, according to a predetermined criterion configured on the basis of predetermined parameters indicative of the type of vehicle engine.
System according to any one of the preceding claims, comprising memory means (20) for recording data indicative of the recognized operating state, as a function of time and of the geographical location of the vehicle, for the generation of data on the use of the vehicle.
System according to Claim 12, comprising communication means adapted to access the said memory means (20) and supply the said data on the use of the vehicle to the outside.
System according to any one of the preceding claims, in which the said processing and recognition means (10) include a microprocessor unit adapted to execute a computer program or set of programs.
System according to any one of Claims 1 to 13, in which the said processing and recognition means (10) include a wired computational logic.
System according to any one of Claims 1 to 13, in which the said processing and recognition means (10) include an applications - specific integrated circuit (ASIC).
System according to any one of the preceding claims, in which the said processing and recognition means (10) are arranged for the self-leaming of a plurality of reference models (48).
System according to any one of Claims 1 to 16, in which the said processing and recognition means (10) are arranged for the guided learning of a plurality of reference models (48).
System according to any one of the preceding claims, in which the said processing and recognition means (10) comprise timer means arranged for providing a periodic wake-up for performing an operation of recognizing the operating state of a vehicle engine.
Method for recognizing the operating state of a vehicle engine, comprising
i) the detection of an acoustic noise signal propagated in a compartment of the vehicle;
ii) the determination of data indicative of the frequency component of the signal and data indicative of the energy of the signal;
iii) the generation of at least one descriptor vector of the detected acoustic signal, comprising the said data indicative of the frequency components of the signal and of the energy of the signal;
iv) the comparison of the said descriptor vector of the detected acoustic signal with a plurality of reference descriptor vectors indicative of known operating conditions, acquired in a learning phase; and
v) the classification of the said detected acoustic signal according to the distance between the descriptor vector of the detected acoustic signal and the reference descriptor vector, so as to discriminate between a first class of signals indicative of an operating condition of the vehicle engine and a second class of signals indicative of an inoperative condition of the vehicle engine.
Method according to Claim 20, additionally comprising the determination of data indicative of a phase shift between at least one pair of acoustic signals received from points displaced in the compartment of the vehicle, and the generation of a descriptor vector of the detected acoustic signal comprising the said data.
Method according to Claim 20 or 21, comprising:
i) the detection of a vibration noise signal propagated in a compartment of the vehicle;
ii) the determination of data indicative of the frequency component of the signal and data indicative of the energy of the signal;
iii) the generation of at least one descriptor vector of the detected vibration signal, comprising the said data indicative of the frequency components of the signal and of the energy of the signal;
iv) the comparison of the said descriptor vector of the detected vibration signal with a plurality of reference descriptor vectors indicative of known operating conditions, acquired in a learning phase; and
v) the classification of the said detected vibration signal according to the distance between the descriptor vector of the detected vibration signal and the reference descriptor vector, so as to discriminate between a first class of signals indicative of an operating condition of the vehicle engine and a second class of signals indicative of an inoperative condition of the vehicle engine.
Method according to Claim 22, additionally comprising the determination of data indicative of a phase shift between at least one pair of vibration signals received from points displaced in the compartment of the vehicle, and the generation of a descriptor vector of the detected vibration signal comprising the said data.
Method according to any one of Claims 20 to 23, comprising the detection of a condition of movement of the vehicle.
Method according to Claim 24, in which the detection of the condition of movement of the vehicle comprises the measurement of an acceleration of the vehicle.
Method according to Claim 24, in which the detection of the condition of movement of the vehicle comprises the performing of a geographical positioning procedure.
System according to any one of Claims 20 to 26, including a comparison between a descriptor vector of the detected signal and a reference descriptor vector according to a predetermined acceptable range of variance for the magnitude of each component of the vector centred on an expected reference value, two vectors being considered similar if the distance between them is less than a predetermined threshold value.
Method according to any one of Claims 20 to 27, comprising the estimation of the current engine rotation speed by processing the detected signals according to a predetermined criterion configured on the basis of predetermined parameters indicative of the type of vehicle engine.
Method according to any one of Claims 20 to 28, comprising the recording of the data indicative of the recognized operating state, as a function of time and of the geographical location of the vehicle, and the generation of data on the use of the vehicle.
Method according to Claim 29, comprising the communication of the said data on the use of the vehicle to the outside.
Method according to any one of Claims 20 to 30, comprising a timed cyclic execution of operations i)-v).
System for managing vehicle access to a restricted traffic area, comprising means of monitoring the passage of vehicles through points of access to the said areas and means of pricing and charging for the said access,characterized in that it includes means for acquiring data indicative of an operating or inoperative condition of a vehicle recognized by the method claimed in Claims 20 to 31, whereby the said pricing and charging means are arranged for calculating a price and making a charge on the basis of the said data on the operating conditions of the vehicle.