WO2011072879A2

WO2011072879A2 - Failure prediction system in railway networks

Info

Publication number: WO2011072879A2
Application number: PCT/EP2010/007795
Authority: WO
Inventors: Josep Girbau Amat; Manuel Frigola Bourlon; Marc Gispert Segura; Ezio Cappellino
Original assignee: Universitat Politecnica De Catalunya
Priority date: 2009-12-18
Filing date: 2010-12-20
Publication date: 2011-06-23
Also published as: WO2011072879A3; ES2374465A1; ES2374465B1

Abstract

The invention relates to a failure prediction system in networks of vehicles which travel on rails, said system uses several types of sensors which capture data relating to various parameters which are processed and sent to carry out an analysis thereof and determine and predict possible failures in the network.

Description

FAILURE PREDICTION SYSTEM IN RAILWAY NETWORKS

Object of the Invention

The present invention relates to the field of railway networks, more specifically to those in which vehicles with an electric motor move on the rails thereof.

The object of the invention consists of a system which allows predicting and determining failures in the vehicles forming part of said railway networks.

Background of the Invention

The birth of the metro, that is, of the electric train which moves using its own exclusive circuit in an urban context, dates back to 1890, the year in which the first line was opened in London.

During more than a century, these systems have undergone an important transformation which has turned the current metro into a showcase of the innovations in the industrial and technological field, in the operation thereof and in the services offered to the client.

Most metropolitan railway trains are electrical "multiple units" with lengths of three to over ten cars. The electricity for the electric motorisations is supplied by a third rail or catenary, in some trains a linear motor is used.

Railway transport networks have monitoring and control systems such as the ones known as Commit, the objective of which is to optimise the maintenance of the installations of the Madrid Metro, reducing the number of incidents, the resolution time thereof and the users affected. The equipment for signalling, communications, energy, station control (ticket machines, turnstile lines) escalators, elevators, video surveillance, travel information systems, etc., are controlled from the control centre. At present, Commit monitors only thirty percent of the stations in the network.

In the event of an incident occurring, the staff of the centre receive a warning call and together with the person who announces it, they try to determine symptoms and causes of the failure. If they do not manage to resolve it, Commit technicians can access and run maintenance operations remotely from their workplace. As a result of this, the number of staff displacements to the place of the incident has decreased by five percent and said reduction is expected to reach fifteen percent within a period of two years.

Equipment monitoring is also carried out remotely from Commit by means of capturing warnings and processing information of the status of the facilities, whereby the prevention and early detection of service losses are achieved. The system is capable of learning from previous experiences in resolving incidents, for which it manages the acquired knowledge and it is fed back from the failure analysis and operation variables of the facilities.

It is therefore necessary to develop similar systems which provide greater possibilities thereto and in turn allow the use thereof in those areas of the network that are remote or are difficult to access, thereby preventing the need for a technician or operators who have to travel to said area; such as those areas of a metro network, in which there are point motors subject to substantial loads, vibrations, low temperatures, humidity, often even floods, and various added drawbacks. Said motors are located in the metro tunnel networks, due to their geographical location the technicians that check them are often required to walk over a kilometre in order to reach them. Their low accessibility and the conditions in which they work, combined with the fact that this involves metro user safety, make new solutions necessary.

Description of the Invention

The system object of the invention proposes a breakdown prevention system by means of a statistical analysis and monitoring of different electrical and physical parameters of the electric motors of the railways of the railway network or metro which allows solving the problems set forth above.

Its application is focused on networks of motors which, due to the number and/or geographical location thereof, complicate their manual monitoring. Furthermore, motors of this type are motors which, due to their application or the means in which they work, require a high reliability with regards to safety of the industrial processes which they undertake or with regards to the safety of the staff who work with them.

The system object of the invention allows predicting breakdowns which affect metro circulation, also minimising stoppages during in-service hours, moving repair tasks or emergency replacements which are performed at random hours to out-of service hours.

The system object of the invention is formed by two main modules, a capture module and a communications concentrator module, and a third analysis module.

The first module is a data capture module, in fact several data capture modules, which carries out the function of capturing data from the motors and their environment, and by means of a communications protocol based on the CAN protocol, it sends information outside the motor.

The second module consists of a communications concentrator module which collects information from all the data capture modules connected to the motors which are connected thereto and compresses it into minimised data packets which it transmits through a communications network, such as a local or Internet network for example, to a server with a database.

The third module is an analysis module comprising a database and a program which analyses and manages said data in order to, in another human interface application, provide the determined information of all the motors making up the network to be monitored.

The data acquisition module is divided into two parts physically separated from one another, one is exclusively intended for the intelligence and memory of the module itself as well as for the management of the communications, whereas the other one is a data acquisition submodule of the system in which it is installed by means of various types of: electrical, physical and mechanical sensors. These two modules are separated and are joined in the assembly, such that they work independently, any of the two being able to be replaced to acquire different functionalities, without the replacement of one of the parts affecting the system as a whole.

Said data capture module is a non-intrusive plug and play-type element which, by means of a connection in parallel to the power supply of the electric motor, can acquire different data and parameters which are used to be able to determine if said electric motor suffers a malfunction or is in the process of deterioration, thus being able to predict a breakdown. The capture module comprises sensors which allow collecting data related to parameters such as:

Voltage (AC and DC): This parameter is important in order to determine the energy that the motor consumes and, in general, the impedance of the motor windings. Since no matter how stable the power supply of the electric network is, it always has small changes both in amplitude and in phase, it is essential to know at all times in what form this reaches the motor, thus assuring the maximum information possible. This parameter is measured in parallel with the motor, and both the sensors and the data capture module are galvanically isolated, whereby they are not intrusive for the motor nor do they imply an additional consumption

Current (AC/DC): The data capture module acquires data related to the current which passes through the motor windings, thus being able, together with voltage measurement, to know the impedance of the latter, how long they consume energy, the energy consumed, and stopping and starting currents. Another essential factor is the frequency analysis of said current. Said current can be measured through a Hall effect sensor, i.e., measured by means of the magnetic field generated by the current that passes through the power cable of the motor, whereby again this implies no intrusion for the latter.

Vibration in high, medium and low frequency: Since the data capture module is coupled integrally to the case of the motor, every vibration it has will be transmitted to the data capture module. The resulting vibration in the direction parallel to the shaft of the motor is particularly critical, such that this vibration is the most important vibration that the data capture module collects, this datum can be captured by means of accelerometers.

Internal temperature of the device and external temperature: By means of two temperature sensors, the data capture module collects the environmental temperature and the temperature in the power supplies of the motor. This datum is essential to correctly calibrate other information such as voltages and currents, as well as being useful to distinguish between environmental temperature changes and those changes generated by the motor.

Check voltages: This parameter is very specific for metro motors, which have a check circuit which, by means of direct voltages indicate the position the motor is in and if it has reached the end of stroke. By means of these voltages, the system is capable of monitoring the status of the wiring and the stroke times. These voltages are acquired in the same way as those of the power supply, i.e. galvanically isolated.

Floods: By means of a flood sensor, the data capture module is capable of warning the system when water enters the case of the motor, protecting itself by means of its shutdown and further warning in advance of a possible short-circuit before this occurs.

The data capture module allows the reprogramming of its firmware (set of programming instructions for specific purposes recorded in a non-volatile memory which establishes the lowest level logic which controls the electronic circuits of any type of device and the functionality thereof) which can be performed from a remote point, such as a control room. The same process can make said firmware change for one or for all the data capture modules connected to the network regardless of the distance between devices.

This reprogramming is useful in the event of changes of the environment or of the evaluation principle of the parameters analysed, such as:

Changes due to climatic conditions: some parameters can be more influential when the climatic characteristics change.

Changes of the loads of the motors throughout their use: according to the application of the monitored motors, the load applied can vary and consequently the characteristics of some parameters can in turn vary.

Changes of the resolution of a parameter it can be detected that useful information of a parameter can is limited in a restricted time range or in a determined frequency component and consequently the resolution of the measurement of this value can be changed.

Through this reprogramming, it is possible to change some intrinsic characteristics of the data capture module without physically accessing the location thereof. For example, the speed of communication with the communications concentrator module, or the priority of its messages in the network can be modified.

The data capture modules have special implemented functions, not only for monitoring or acquiring electrical and mechanical data of a motor and the system in which it is comprised, but, due to the architecture of the capture module itself, they can also have action functions on this motor and system. It can thus be asserted that each of them is a data capture module which has the possibility of behaving as a preventive action. Each capture module can have, depending on the case, a dual function:

Data Capture: The data capture module carries out the function of receiving electrical and mechanical data of the motor and the system in which it works by means of its instructions, implemented in the firmware described above, performs a previous analysis, calculations and previous packing of this information.

Action in the system of the motor: The data capture module can incorporate functions which act directly on the system in which the motor works, thus being able to apply a control on the operation thereof. A clear example can be the automatic disconnection of the motor by the data capture module if the latter detects the presence of water within the area of the system.

The communications concentrator module is the module in charge of concentrating the communications of all the data capture modules which are connected thereto and consists of:

Motherboard: The main element of this module is in charge of supplying energy to the data capture modules by means of a power supply source. Furthermore, it also preferably contains 8 CAN (Control Area Network) data buses as connections for the data capture modules and 1 CAN data bus for the management of internal information defined between the 8 slave boards and the master board. It is also in charge of supplying and distributing energy to the 9 boards (8 slave, 1 master). The motherboard has a display which connects to the master board and the different slaves. Information is shown on this display concerning how many data capture modules are connected to each slave board and if information, activity, etc., is being received. Lastly, it also has communication output elements, such as RJ45-type connections, for example, in turn connected to the master board. The motherboard also has fuses to protect the power supply line in the event of the fuse being blown, an indicator would shut off. Slave board. This board receives information from the data bus connected to the different data capture modules, up to 8 of them along a kilometre and a half of bus. It is connected to the internal data bus to communicate with the master board and decides whether or not the motherboard must supply energy to the data capture modules which are connected thereto.

Master board. The master board is solely connected to the Control Area Network internal bus of the motherboard, and it condenses the information from the 8 slave boards and transmits it by an SPI bus (Serial Peripheral Interface Bus) to the motherboard, which in turn transforms the data packets of this bus to the Ethernet standard. It also has a temperature control.

The main function of the communications concentrator module is that of concentrating the communications of the data capture modules and then efficiently sending them, compressing them into few data packets, by Internet or a local area network with connection to a server which contains the database.

One of the secondary functions of the communications concentrator module is that of administering the power supply or energy supply to the data capture modules, i.e., the latter do not need an extra power supply, since it is supplied by the same communications cable from the communications concentrator module, such that the communications concentrator module can supply energy to the underlying data capture modules or not, as preferred. For example, it can cut off the power supply of one or more data capture modules which are underwater due to floods.

The data capture module may not necessarily require power supply from the motor itself, each data capture module is fed from the communications concentrator modules. This power supply is not a simple power supply through a power supply source or from the electric network, but it is a result of a power stage and control stage.

The communications concentration module manages the control of the power supplies, such that it is capable of interrupting that power supply to automatically reset the data capture modules, also to protect them in the event of floods or malfunctions, etc., such that the system object of the invention has the following possibilities with regards to power supply of the data capture modules:

Autonomous power supply of the data capture modules: They do not require a connection for their power supply nor a battery next to the motor or system in which they are installed. Management of the power supply from the communications concentrator: The concentrator module has different power and control stages capable of supplying energy to the data capture modules using the same cable which is used to perform the data communications. Possibility of interrupting the emergency power supply and resetting the data capture modules: There are cases in which the data capture module may need to be disconnected in order to protect itself, the motor or the system in which it is installed. An example of the first would be after a flood. The second case for example could occur if a data capture module started to operate anomalously, saturating the communications line, which could damage the communications concentrator, and the latter could decide to shut it down in order to not compromise the entire system. The third case could occur if it had installed therein a system for the automatic disconnection of an electric motor by means of a data capture module, if the latter disconnected the motor without there having been any repeated warning signal or anomalous situation, which would indicate a malfunction of the hardware of the data capture module, the system would disconnect that data capture module, making the motor operate without interruption

Another function of the communications concentrator module is that of reprogramming all the data capture modules connected thereto. By operating from the user interface of the server of the analysis module, the communications concentrator module can change the software/firmware of the data capture modules remotely, thus allowing reconfigurations of the system from the control room.

The communications concentrator module is designed to be connected to the electric network and can be encapsulated in a standardised commercial case for a rack-type cabinet; in a preferred embodiment it has 8 input/output connections for connecting to the data capture modules, it also has network connections, such as RJ45 Ethernet network connections.

The communication between the data capture modules and the communications concentrator module is performed on the CAN low-level protocol. One of the main advantages of said protocol lies in the possibility of covering large distances using a small number of cables for that purpose.

Given that this protocol only assures that one data packet is received, it is correctly received exactly how it was sent; i.e., when using this protocol, there are no assurances that the packets are received in the same order as they are sent nor that all the packets will be received, to that end a layer has been implemented which is added to the ISO/OSI communications stack at a higher level. Despite the existence of high-level protocols based on CAN, such as CANOpen, a new, simpler approach has been implemented to reduce operating loads of the system elements. This protocol is based on timed data requests which assure, through redundancy, the success of the communication without losses and maintaining the sending order.

The system is completed with the analysis module, which is based on a client-server architecture.

On one hand the server is in charge of collecting the data from the previous modules and saving them in a database. In order to receive data, a UDP (User Datagram Protocol) server is used which receives packets from the communications concentrator modules connected in the same communication network. In parallel, the server of the analysis module is in charge of performing the necessary statistical analysis for predicting possible failures in the operation of a motor.

On the other hand, wherever the client is, it can view and control the status of the network of motors in which the system is located, provided it is connected to the same network as the server of the analysis module.

The server of the analysis module is run in a remote server in which the database can also be stored, which has to remain continually active in order not to lose data which could be important for the analysis. Another option is to use "mirror techniques" in which the data are duplicated in different locations.

The client can be run from any computer without important technical characteristics.

The core of the analysis module resides in its database, in which the data of all the network of motors are stored.

In order to assure its consistency, only writing access to the server of the analysis module is allowed, whereas clients can only gain reading access to view the necessary information for predictive maintenance tasks.

The server of the analysis module is made up of four main parts:

The UDP connection.

- The statistical analysis processing means.

The database located in one or several digital storage media, such as solid memories, hard drives, optical supports, etc...

The configuration interface.

The UDP connection is in charge of receiving the data from the communications concentrator modules and storing them in the database.

A process is run in parallel which is in charge of performing a statistical analysis based on the mean value of the data and the standard deviation thereof over time.

The server of the analysis module receives the characteristic values of the data of the motors: peak current, effective voltage, movement time, amplitude of the vibration in low, medium and high power, etc. Errors are predicted and warnings are generated from these data.

Each signal acquired in the data capture module, through the rest of the system, is processed again generating a reliability percentage, in which 100% indicates that the value obtained from the data capture module is within the margins of normality and 0% indicates that it is completely outside these margins. If the percentage is below the threshold, the warning associated to the value which has given rise to that percentage is generated.

Once all the percentages indicative of the status of each signal acquired from the data capture module have been calculated, a value which describes the global status of the motor is calculated. This calculation consists of the minimum selection of all the weighted signals according to their relevance.

The calculation of the reliability percentage of a specific signal focuses on the analysis of the deviation of the last sample according to the record of previous samples. A value of these data which does not vary over time for a motor is indicative of its correct operation. When any of these values begins to be altered, the analyser generates warnings associated to this unexpected variation.

By updating the mean value and its variance every time new values are received, the system learns to discard those data which do not represent a threat to the prediction of errors.

The values of this sample are adjusted automatically based on the interaction with the users of the system, who, during an initial learning phase, will record what incidences have been correctly predicted, and which have occurred without generating warnings. These parameters can also be manually adjusted using the configuration interface in which the basic parameters characterising the analysis, such as the variance threshold of the values, or the number of evaluated data, can be adjusted according to the needs of the system.

In order to perform these changes said configuration interface is available, only accessible to enabled users, which allows varying these values at any time without the need of restarting the server.

The client of the analysis module allows viewing the status of the motors and of the devices through a computer interface. This application can be run from any point of the network and consistently accesses the database of the analysis module in order to extract the information necessary for the predictive maintenance tasks.

In the event of anomalies in the typical values of a motor, different types of warnings are activated which can be remotely managed by enabled users.

Description of the Drawings

To complement the description which is being made and for the purpose of aiding to better understand the features of the invention according to a preferred practical embodiment thereof, a set of drawings is attached as an integral part of said description, in which the following has been depicted with an illustrative and non-limiting character:

Figure 1 shows a schematic view of the wiring of the data capture module.

Figure 2 shows a view of the components of the communications concentrator module.

Figure 3 shows a diagram of the system and its modules.

Figure 4 shows a flowchart of the statistical percentage analysis which the statistical analysis module of the system carries out.

Preferred Embodiment of the Invention

A preferred embodiment of the system (1 ) object of this invention is described below in view of the figures.

In a preferred embodiment of the invention, the system (1) applies to railway point motors (5) of the metro network. Due to the location of these motors (5) and their critical nature for the correct operation of the metro network, is of vital importance to be able to obtain a preventive maintenance which allows performing maintenance operations before malfunctions occur which can cause a stoppage of a line or the entire network.

By means of a data capture module (2), which has several sensors, such as electrical voltage, vibration, current or flood sensors, and which is connected in parallel to the motor (5), as seen in Figure 1 , data relating to the environment of the motor (5) are captured and from these data an anomalous situation is determined by means of the flood sensor of the data capture module (2) relating to the flooding of the motor (5) due to infiltrations of water in tunnels; in turn, by means of the vibrations sensor of the data capture module (2), anomalous vibrations are detected in the motor (5) due to mechanical problems of the motor (5) or a weakness in the anchoring to the track.

Said data collected by the data capture module (2) are sent to a communications concentrator module (3) which collects the data from the data capture module (2), and it supplies energy to the latter by means of a power supply source (9) through a configuration such as the one seen in Figure 2, which has slave boards (7) connected to an internal CAN bus of a motherboard (6) to which there is connected a master board (8) which is in charge of compressing said data and transmitting them in the form of data packets to the motherboard (6) through an SPI bus. Said motherboard (6) is in charge of sending them, once they have been converted to the Ethernet standard, to an analysis module (4) through a local communications network, as seen in Figure 3.

The set of collected data is stored in a database in order to use them in a statistical analysis which is performed by analysis processing means comprised in the analysis module (4). Said analysis is based on previously established parameters and on a statistical percentage calculation, in which 100% indicates a good status of the motor (5), said calculation is depicted in the flowchart of Figure 4 to distinguish between a correct operation or malfunction and to even determine the time prior to a breakdown.

Claims

1. A failure prediction system (1 ) in railway networks, characterised in that it comprises:

- at least one data capture module (2) adapted to be electrically connected in parallel to an electric motor (5) of a railway vehicle to capture data from the electric motor (5) and control said electric motor (5), comprising sensors and means of acting on said electric motor (5), at least one communications concentrator module (3) connected to the data capture module (2) in charge of collecting and processing the data captured by said data capture modules (2) and controlling the energy supply thereof, and

- an analysis module (4) which is located in a remote location comprising programmable devices and digital data storage devices in charge of analysing and storing the data sent through a communications network from the communications concentrator module (3) .

2. The system (1 ) according to claim 1 , wherein the sensors of the data capture module (2) are selected from among the following: temperature sensors, voltage sensors, current sensors, vibration sensors, flood sensors and check voltage sensors.

3. The system (1) according to claim 1 or 2, wherein the data capture module (2) comprises a power cable connected to the communications concentrator module (3) in charge of supplying energy for its operation by means of a power supply source (9).

4. The system (1) according to claim 1 or 2, wherein the data capture module (2) comprises autonomous power supply in charge of supplying energy for its operation.

5. The system (1 ) according to any one of the previous claims, characterised in that the communications concentrator module (3) comprises:

a motherboard (6) comprising several CAN data buses to communicate with the capture modules and an internal CAN bus,

- at least one slave board (7) connected to the internal CAN bus of the motherboard (6) and which is adapted to receive the data from the data capture modules through the CAN data buses, and

a master board (8) connected to the internal CAN bus of the motherboard (6) and which is adapted to compress the data from the slave boards (7) and transmit them in the form of data packets by means of an SPI bus to the motherboard (6) which is in charge of converting said packets to the Ethernet standard.

6. The system (1) according to any one of the previous claims, wherein the communications concentrator module (3) is adapted to compress and pack the data received from the data capture module (2) and to send them to the analysis module (4).

7. The system (1) according to any one of the previous claims, wherein the communications concentrator module (3) is adapted to reprogram the data capture module (2).

8. The system (1) according to any one of the previous claims, wherein the analysis module (4) comprises:

a communication interface,

- a UDP connection in charge of receiving the data from the communications concentrator module (3) ,

a database located in digital storage media in charge of storing data from the UDP connection, and

an analysis processing means in charge of performing a statistical percentage analysis of the data from the data capture modules (2).

9. The system (1) according to the previous claim, wherein the analysis is performed by means of a programmable device in charge of running instructions stored in the digital data storage media.

10. The system (1) according to any of the previous claims, wherein the communications concentrator module (3) additionally comprises a display in charge of supplying information relating to the operation of the boards (6, 7,8).