RU2569216C2 - Method of control over servicing and repair of railway rolling stock and system to this end - Google Patents

Abstract

FIELD: physics, control.

SUBSTANCE: invention relates to railway ACS. Claimed process comprises control steps. Incident control step includes collecting and classifying of incidents revealed by diagnostics. Problem control step includes processing of classified incidents by mathematical statistics and factor analysis methods and, proceeding from the data of augmentable database, defining of troubleshooting process at correcting measures. Servicing control step includes the analysis of key symptoms of servicing quality and repair, their interrelation with revealed incidents and/or problems and execution of correlation analysis of the entire accumulated data base to update controlled parameters, order of their collection and/or measurement intervals. Claimed system comprises the server with data I/O system and software, electronic terminals and measuring hardware including locomotive onboard control systems, stationary and portable diagnostics systems.

EFFECT: higher reliability of railway equipment.

4 cl, 3 dwg

Description

The invention relates to the field of reliability management systems for railway transport, and in particular to systems for research, analysis and adjustment of maintenance and repair of railway transport.

The prior art method and system for controlling the transfer of electronic data between the service diagnostics center and many remote mobile devices, in which the electronic data stored in the database reflect the operational qualities of these devices and are used to determine system malfunctions and errors. The known method allows you to trace the history of each case and use its analysis during subsequent operation to detect the presence of relevant potential faults in other mobile devices (see patent US 7051044, CL B61L 27/00, publ. 23.05.2006).

A disadvantage of the known solutions is the lack of systematic processing of information that would reliably predict the operability of railway transport and offer timely corrective measures.

The prior art also knows a method for maintenance and repair of traction rolling stock (TPS) of a railway vehicle according to its technical condition, including diagnosing rolling stock with stationary bench or portable diagnostic systems at the stages of maintenance and current repair and predicting its technical condition after maintenance and maintenance (see patent RU 2487023, CL G01D 21/00, publ. 07/10/2013).

The main disadvantage of this method is that it considers only the diagnosis of individual nodes of the locomotive, and not the locomotive as a whole, while the serviceability of individual nodes of the system does not guarantee its operability as a whole.

Closest to the technical nature of the claimed invention is a method of managing maintenance and repair of railway transport, including collecting data for diagnosing the current technical condition of the equipment, processing and corrective actions based on data on the results of such processing (see patent RU 2357215, class G01D 21/00, published May 27, 2009). From the same source, a control system for maintenance and repair of railway transport is known, including at least one server with an information input-output system on which software is installed, and measuring equipment that transmits data to the specified server.

The disadvantages of the known technical solution is that only pre-emergency conditions of traction rolling stock (TPS) are considered and only stationary bench and portable diagnostic systems are used. In the well-known solution, only three technical conditions of the TPS are identified: operational, limited-functioning and pre-emergency, not giving a complete picture of the current state of the traction rolling stock of railway transport.

The objective of the invention is to remedy these disadvantages. The technical result is to increase the reliability of the traction rolling stock of railway transport.

In terms of the method, the problem is solved, and the technical result is achieved by the fact that the method of controlling the maintenance and repair of traction rolling stock of railway transport, including collecting diagnostic data on the current technical condition of the equipment, processing them and taking corrective actions based on data on the results of such processing, includes the incident management stage, the problem management stage and the service management stage, and at the incident management stage I carry out collection and classification of incidents detected according to diagnostics, which use data from on-board locomotive control systems, from stationary and portable technical diagnostics systems, as well as data on ongoing maintenance and repair, at the stage of problem management, they process classified incidents using methods mathematical statistics and factor analysis, in particular correlation analysis, identify the causes of incidents and based on data the updated information base, they determine the preferred method for eliminating these causes during corrective actions, and at the service management stage, analyze key indicators of the quality of maintenance and repair, their relationship with identified incidents and / or problems, conduct a correlation analysis of the entire accumulated database and, if necessary, change the parameters controlled during the control process, the procedure for their collection and / or the frequency of measurements.

In terms of the device, the task is solved, and the technical result is achieved by the fact that the control system for maintenance and repair of traction rolling stock of railway transport includes at least one server with an information input-output system on which the software is installed, and measuring equipment, transmitting data to the specified server, and the specified at least one server includes interconnected incident management unit, problem management unit, unit service management, and the incident management unit collects and classifies incidents detected by the diagnostic data obtained from measuring equipment, including on-board locomotive control systems, stationary and portable technical diagnostic systems, as well as data on ongoing maintenance and repair recorded in electronic the terminals associated with the specified incident management unit, the problem management unit is equipped with a replenished information base and flnn with the possibility of processing classified incidents using methods of mathematical statistics and factor analysis, in particular correlation analysis, identifying the causes of incidents and determining the preferred method for eliminating these causes during corrective actions, and the service control unit is able to analyze key quality indicators of technical maintenance and repair, their relationship with identified incidents and / or problems, correlate analysis of the entire accumulated database and, if necessary, change the parameters controlled during the management process, the procedure for their collection and / or frequency of measurement. The information base of the problem management unit preferably includes regulatory documents and is configured to be automatically replenished via the Internet. The information base of the problem management unit can be equipped with an interface for publishing identified problems requiring solution on the Internet.

Figure 1 presents the algorithm for controlling the reliability of locomotives for the PDCA cycle;

figure 2 - algorithm of the proposed method of managing maintenance and repair of railway transport;

figure 3 is a schematic diagram of the proposed control system for maintenance and repair of railway transport.

The fundamental principle of technological process management is the principle of continuous improvement, which is fixed in a number of international ISO standards (9000-9004, 20,000, etc.), and is included in the concepts, standards, and methods of managing the quality, reliability, and safety of Russian Railways.

“Quality” and “Reliability” as objects of continuous management were introduced in the 30s of the 20th century by Walter Andrew Shewhart as part of the development of “statistical quality control” systems that recognize the inevitability of “defects” and the need to gradually reduce them quantity by eliminating the root causes. Edward Deming (William Edwards Deming) developed the ideas of W. Shuhart and proposed a systematic approach, known as the "Deming Cycle" (PDCA cycle: Plan, Do, Check, Act) - plan, execute, analyze the result, conduct corrective actions.

The PDCA cycle is the obligation of feedbacks that represent the methodological basis of another science - Cybernetics (in technology - the theory of automatic control).

The principle of continuous improvement is also implemented when managing the reliability of the locomotive (Fig. 1). As a result of the operation of the TPS, an initial database of failures and malfunctions is formed. Periodically, a factor analysis is performed in the depot, the data are summarized. At each control level, corrective actions are formed aimed at improving the reliability of the locomotive according to the PDCA cycle. Analytical reports are carried out including in universities and research institutes.

Existing railway transport reliability management systems have a number of features that create problems for management.

1. Failures of technical equipment are considered as failures of the transportation process: service interruption, train being late or delayed, traffic safety violation, etc. This is convenient for controlling the transport complex, but there is no interpretation of the failure as a malfunction.

2. The classifier of dangerous events of railway transport used in the safety and reliability management system (Order of the Ministry of Transport of the Russian Federation No. 163 of December 25, 2006) contains events that are dependent on each other, which complicates the use of the mathematical apparatus of the theory of reliability. For example, the events “Traffic light with a prohibiting signal”, “Collision”, “Descent”, etc. can occur simultaneously.

3. Different consequences of events, their significant heterogeneity and incomparability in terms of consequences (for example, train delay and descent) when taking them into account as “failure” is also inconvenient when managing reliability.

4. Problems of using statistical management methods due to their relatively low intensity.

5. The absence of division of failures into malfunction, inoperability and improper functioning reduces the reliability of factor analysis.

6. The subjectivity of failure accounting due to manual accounting is high. The procedure for assigning responsibility for farms is not debugged. With a different system of employee motivation, the psychological restrictions on entering reliable information substantially change. The data is disparate.

7. Formally, the number of service failures will increase, taking into account the actual condition of the locomotive and unscheduled preventive repairs.

8. One and the same event can be recorded as a failure or not depending on their performance on planned or unplanned types of repairs.

9. Failures of technical means do not directly affect the volume of transportation and income, therefore, at Russian Railways, the effect of technical means is recorded in lost train hours (form DO-14VTs). This indicator has subjectivity - the indicator “Lost g * km gross” could be more reliable. In the locomotive complex, the parameters “Availability”, “Simple for repair”, “Need for component parts and materials”, “Labor costs”, “General expenses of Russian Railways for maintenance and TP of locomotives”, etc. are more indicative for reliability management.

10. The frequency of analysis has annual and shorter periods. The analysis shows that the response of the technical complex to corrective measures is 2.5-3 years. A resource increase has even longer periods of observation time.

11. There is no statistical validation of the source data. For example, there is no check for single-mode processes. In the locomotive complex, failure rate, influencing factors (train mass, speed, power, etc.) and other parameters are incorrectly taken into account.

12. Factor analysis is not automated. The level of training of specialists does not allow the use of mathematical analysis in practice in the absence of automation.

Thus, to effectively manage the reliability of rail transport, a number of problems need to be addressed.

With the control of occupational injuries, the methodology of the American scientist G. Heinrich (Herbert William Heinrich) became classical. After analyzing 550 thousand accidents, G. Heinrich deduced the pattern ("Heinrich’s Pyramid"), according to which 29-30 accidents with less severe consequences and 300-330 "non-significant" incidents occur in one serious injury. It is proposed to analyze not single cases of injuries, but to work with representative statistics of the “base of the pyramid” (the method is described in 1931 in the book "Industrial Accident Prevention: A Scientific Approach").

The ratio of 1: 30: 300 is not constant. For example, DuPont, one of the leaders in industrial safety, uses a different proportion and number of pyramid levels. According to the DuPont methodology, it can be argued that 10 thousand cases when a person stepped on the rail head will lead to 1 thousand cases of falling, 100 cases of injuries with 50 cases of temporary disability and one permanent. Heinrich’s pyramid was also used in Russian Railways for managing train safety.

In the context of railway transport, the Heinrich pyramid principle indicates that the operation of a locomotive leads to incidents, which leads to malfunctions with the subsequent possible occurrence of failures in the transportation process.

The Heinrich Pyramids are similar to the Five Whys (5W) methodology proposed by Sakichi Toyoda, one of the founders of the Toyota Production System (TPS), better known as Lean Production and Lean Production: When investigating an event, you need to answer the question “Why?” as many times as you need to find out the root cause of the incident. Obviously, the number of questions corresponds to the number of levels of the Heinrich pyramid.

There are two main differences between the Five Why? Method and Heinrich’s pyramids: the 5W method does not have a numerical relationship of levels, but there is a tree-like (hierarchical) representation of cause-effect relationships - Ishikawa Diagram (Fish Skeleton, Fishbone) ) The Ishikawa diagram in response to the question “Why did the locomotive fail?” Would have the following “Thick bones”:

- The design flaws of the locomotive and its components and equipment;

- Poor production or overhaul;

- Poor maintenance and repair in a depot;

- Violation of operating modes;

- Increased wear on equipment and components.

The second “Why?” Will have the answers:

- Human factor;

- Lack of technological equipment;

- Non-compliance with established rules and regulations;

- External causes that caused the previous reasons;

- The tradition is to work like that.

To the third “Why?” For the answer “Human factor” answers are possible:

- Low qualification;

- Bad motivation;

- Shortage of personnel;

- Indiscipline;

- A disease.

The hierarchy of questions and answers can be continued.

To date, the most effective element of the decision support system, as well as a convenient method for identifying unreliable initial statistical data and cases of violation of the railway operation technology, is a correlation analysis.

Combining the advantages of Heinrich’s pyramids, “5W” methodologies and Ishikawa diagrams can be done using correlation analysis: the ratio 30: 300 corresponds to the correlation coefficient r = 0.1 when considering the dependence as binary: 0 - there was no event, 1 - was. In more complex cases, you need to choose the right metrics. For example, between the skill level and the number of failures: if a locksmith of the 6th category does not have a marriage, the 5th - 10%, the 4th - 20%, the 3rd - 30%, and the 2nd - 50 %, then r = -0.46, which is much higher than the previous example. Another example, if every 5 thousand km of mileage leads to an additional failure, then r> 0.98 (depending on the method of recording information).

Thus, for each branch of the laws, metrics should be found that allow for an adequate correlation analysis.

Heinrich’s pyramid methodology is complemented by the international service standard ISO 20000 and ITIL, a library of recommendations (Best Practice), consisting of 10 books. Books - “Service Support” (Service Support) and “Service delivery” (Service Delivery) - are combined in the general section “Technical support” (IT Service Management - ITSM). ITSM is being successfully implemented at Russian Railways in communications and information technology farms.

According to ITSM, 10 processes should be implemented in the service process. We distinguish three of them, which are essentially three levels of feedback: incident management, problem management, and service management.

The Incidents Management process, which implements the technical assistance functions, is aimed at quickly restoring the system by eliminating incidents - any situations other than normal operation. The objective of the process is to minimize service interruption times. The transition from the concept of “Failure” to the concept of “Incident” is important. For example, the driver’s comments in the logbook “Inrush current in traction” and “Load shedding” are an incident, but not necessarily a failure. In addition, incidents of operating modes by the drivers, unacceptable external conditions, etc. also fall into incidents. In fact, there is a transition from the top of the Heinrich pyramid to its base.

The “Problems Management” process aims to reduce the number of incidents through a study of the root causes. Includes factor and statistical analysis, identification of trends and known errors.

The Service Level Management process allows you to install, monitor and manage the level of customer service. The process defines measurable goals for Service levels, allowing you to justify obligations under Service Level Agreements (SLAs).

It is necessary to automate control in all three reliability management loops. First of all, in terms of fixing the incident. Currently, the railway management systems are equipped with the following automated systems.

The first automated source of information is automated workstations (AWS) for decrypting diagnostic data of onboard microprocessor-based locomotive control systems (AWS MSU). To automate the transmission of diagnostic messages, it is necessary to finalize the automated workstation of the ISU, primarily in terms of automating the search for incidents and the transmission of diagnostic messages. There are several types of ISUs: drive control systems (ISUD, ISUE, MPSU, ISU-TP, ISU-TE, etc.), fuel accounting systems (APK “Bort”, RPDA-T, ASK VIS, ASK VNIKTI, etc.) and Automotive Driving Systems (USAVP). In the process of locomotive control, ISUs interrogate sensors, process and save information, which is then used as diagnostic information: ISU data is read (using flash-cards, a laptop, via Wi-Fi and / or GPRS (GSM)) and analyzed on specialized automated workplaces (AWP MSU). The use of diagnostic data of local government significantly increases the amount of diagnostic information. As a result, a more complete understanding of the technical condition of the locomotive, it becomes possible to fix the precautionary state in real operating conditions, which is not always possible to simulate in stationary conditions, and most importantly, it becomes possible to control the operation modes of locomotives.

The second source of diagnostic messages is portable and stationary automated systems for technical diagnostics (ASTD): automated workstations of rheostatic and equipment testing stations, portable systems of the Doctor and Eye series, vibration testing systems Forecast and Vector, etc. If the diagnostic system If it is not automated (for example, non-destructive testing systems), then appropriate automated terminals should be developed.

It is also necessary to use electronic terminals as a source of diagnostic messages for recording data on ongoing maintenance and repair.

Another possible automated source of information is automatic railway transport control systems (ASUZHT): Russian Railways has a number of information systems that record events with a locomotive: mileage, visits to the depot (ASOUP, GUID Ural, ASUT, etc.) failures (KAS ANT), comments of drivers (ACS ZM), events of security breach (ACS NBD) and others. There are decision support systems (URRAN (risk management), SOLARIS, etc.).

To work with the source information and three-circuit reliability control, a unified information management system for monitoring the technical condition of locomotives is required, which automatically detects incidents by automatic diagnostic messages from the systems listed above and provides automated control over the life cycle of incidents and problems. The basic principles of control are integrated into the system by the principle of error protection according to the Barrier, 8D, 5W2H, 5W and all other methods provided for by the corporate standards of Russian Railways, GOST and ISO.

Based on the above data, a method for managing maintenance and repair of railway transport was developed, including collecting diagnostic data on the current technical condition of the equipment, processing them and taking corrective measures based on data on the results of such processing (the corresponding algorithm is presented in Fig. 2).

The proposed method includes:

- stage of incident management;

- stage of problem management;

- stage of service management.

At the incident management stage (in contrast to the accepted failure management, an incident is any situation other than the normal operation of the railway) collect and classify incidents detected by the diagnosis. As the diagnostic data, data are used from the onboard control systems of locomotives, from stationary and portable systems of technical diagnostics, as well as data on the ongoing maintenance and repair and, if necessary, other data.

At the problem management stage, classified incidents are processed using methods of mathematical statistics and factor analysis, in particular, correlation analysis. In the course of problem management, the causes of incidents are identified and the preferred method for eliminating these causes in the course of corrective actions is determined on the basis of data from the updated information base.

At the service management stage, they analyze key indicators of the quality of maintenance and repair, their relationship with identified incidents and / or problems, conduct a correlation analysis of the entire accumulated database and, if necessary, change the parameters controlled during the management process, the procedure for their collection and / or frequency measurements. Thus, the third circuit improves the overall efficiency of railway service by changing key quality indicators (KPI), including included in the quality level agreement (SLA) or other documents that determine the quality of service. Its task is to adjust the service processes themselves.

The proposed method is implemented using a control system for maintenance and repair of railway transport, including one or more servers with an input / output information system on which software is installed, and measuring equipment that transmits data to the specified server (see Fig. 3).

These servers include interconnected incident management unit 1, problem management unit 2, and service control unit 3, controlling the respective above steps.

The incident control unit 1 collects and classifies incidents detected by the diagnostic data received from the measuring equipment and electronic terminals 4, in which data on the ongoing maintenance and repair are recorded. Measuring equipment includes at least on-board control systems for locomotives 5, as well as stationary and portable systems for technical diagnostics 6. In addition, it is possible to connect to the unit 1 other measuring equipment or other automated systems, such as ACMS 7. As a result of life management the incident cycle forms a database.

Problem management unit 2 is configured to process classified incidents using methods of mathematical statistics and factor analysis, in particular correlation analysis, to identify the causes of incidents and determine the preferred method for eliminating these causes during corrective actions. To do this, periodically or upon request (if the incident is not of a type already known), a factor analysis of the information from block 1 is carried out using statistical control methods.

Block 2 is equipped with a replenished information base 8, which contains regulatory documents and requirements of international, national and corporate standards in the areas of reliability management, quality management, lean manufacturing management, etc. Information base 8 is configured to automatically replenish via the Internet. If a typical incident with an unexplained cause is identified, a new “Known Error” is recorded in block 2. Next, a search is made for possible solutions to the problem using a database of possible technical solutions and Internet information. If there are no ways to fix the problem, their cost or other reason, a “known problem” is recorded. The information base 8 is equipped with an interface for publishing identified problems that need to be solved on the Internet in order to attract organizations that have ready-made solutions to this problem or have developments on this issue. If a solution is found, then corrective measures are developed and implemented that close the second control loop.

The service control unit 3 is configured to analyze key indicators of the quality of maintenance and repair and their relationship with identified incidents and / or problems. Block 3 also carries out a correlation analysis of the entire accumulated database and, if necessary, changes the service technology at all three levels of three-loop control by changing the monitored parameters, the order of their collection and / or the frequency of measurement.

The advantages of the proposed solution are due to the following factors. In addition to pre-failure conditions, operational mode violations are considered, which for a feedback system gives a significantly better result. Statistical methods for managing reliability are used: trends in parameter changes, the intensity of failures (including protection trips) are identified, previous statistics are used on the rate of development of the precautionary state in failure. Thus, it becomes possible to carry out maintenance and repair according to the actual state by predicting the remaining life.

The data of the entire life cycle of the incident is analyzed: the time it takes to repair, the time of repair, the time to search for a spare part, etc., the quality of the repair performed, the number of participants in eliminating the incident, and the expenditure of other types of resources. The accumulation of these data allows the further implementation of a three-circuit (and not single-circuit, as in the traditional case) system for managing the reliability of railway transport. The principle of continuous improvement of PDCA is being implemented, which meets the requirements of a number of standards, including ISO 9001.

The result of the implementation of the proposed method and control system for maintenance and repair of railway vehicles is to increase the reliability of locomotives, reduce the downtime of locomotives in repair locomotive depots for repairs, increase the operational availability of locomotives, reduce the cost of operation, maintenance and repair of traction rolling stock.

Claims (4)

1. A method for managing maintenance and repair of traction rolling stock of railway transport, including collecting diagnostic data on the current technical condition of the equipment, processing them and taking corrective actions based on data on the results of such processing, characterized in that it includes an incident management stage, a problem management stage, and a service management stage, and at the incident management stage, the collection and classification of incidents detected by the diagnostic Using data from on-board locomotive control systems, from stationary and portable technical diagnostic systems, as well as data on ongoing maintenance and repair, at the problem management stage, they process classified incidents using methods of mathematical statistics and factor analysis, in particular correlation analysis, identify the causes of incidents and, based on the data of the replenished information base, determine the preferred method for To eliminate these causes during corrective actions, and at the service management stage, analyze key indicators of the quality of maintenance and repair, their relationship with identified incidents and / or problems, conduct a correlation analysis of the entire accumulated database and, if necessary, modify the process control parameters, the procedure for their collection and / or frequency of measurements. 2. The control system for maintenance and repair of traction rolling stock of railway transport, comprising at least one server with an information input-output system on which software is installed, and measuring equipment that transmits data to the specified server, characterized in that at least one server includes interconnected incident management unit, problem management unit and service control unit, the incident management unit collecting and classification of incidents detected by the diagnostic data obtained from measuring equipment, including on-board locomotive control systems, stationary and portable technical diagnostic systems, as well as data on ongoing maintenance and repair recorded in electronic terminals associated with the specified incident control unit, the problem management unit is equipped with a replenished information base and is configured to process classified incidents with using methods of mathematical statistics and factor analysis, in particular correlation analysis, identifying the causes of incidents and determining the preferred method for eliminating these causes during corrective actions, and the service control unit is able to analyze key indicators of the quality of maintenance and repair, their relationship with the identified incidents and / or problems, carry out a correlation analysis of the entire accumulated database and, if necessary, change nyat controlled process control parameters, the order of acquisition and / or frequency measuring. 3. The system according to claim 2, characterized in that the information base of the problem management unit includes regulatory documents and is configured to automatically replenish through the Internet. 4. The system according to claim 3, characterized in that the information base of the problem management unit is equipped with an interface for publishing identified problems requiring solution on the Internet.

Images