JP2008171026A - Maintenance method and maintenance device for system, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve the operating rate of a system during maintenance to reduce total cost. <P>SOLUTION: This maintenance device for the system has: a database 13 storing a type name, a use time and a failure rate of each of constructional apparatuses, and a boundary time of each the constructional apparatus in the system having the constructional apparatuses; an input part 11 inputting the type name and the use time of a failed constructional apparatus in time of post-maintenance or predictive maintenance; a type extraction means 14A extracting the constructional apparatus of the same type name from the constructional apparatuses of the database based on the input type name of the constructional apparatus; an accidental failure type decision means 14B deciding the constructional apparatus having a failure rate transferring to an accidental failure type from failure rates of the extracted constructional apparatuses; a use time decision means 14C deciding whether the use time of the decided constructional apparatus is longer than the boundary time of the constructional apparatus of the same type name or not; and a means 14D deciding the constructional apparatus decided that it is longer as the replaceable constructional apparatus. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a system maintenance method, a maintenance apparatus, and a program for carrying out maintenance operation of various systems used for processing of basic business.

  Various information processing systems, plant systems, and the like can be used as systems used for enterprise core business processing.

  In the conventional classical system maintenance method, in the case of a system with early failure type components whose failure rate decreases with time, periodic preventive maintenance is carried out, and an unfailed state deteriorated at this stage. When a component device comes out, it is replaced with a new component device. However, when an unfailed component device is replaced with a new component device, the failure rate of the component device immediately after the replacement increases, and the system failure increases.

  Therefore, as a result of repeated long-term experiments and research on whether periodic preventive maintenance is an effective maintenance method for improving system reliability, it has been found that it is a factor that increases failures. That is, it has been found that the deterioration of the component equipment including the parts does not depend on the passage of time, and the failure rate of the system is reduced if the classical preventive maintenance is not performed. It has been suggested that an optimal maintenance method should be performed without performing periodic preventive maintenance (Non-Patent Document 1).

In recent years, there are many examples of system maintenance methods that perform post-mortem maintenance and predictive maintenance instead of periodic preventive maintenance based on the above recommendations. Post-maintenance is a maintenance procedure to replace a new component device when a component device failure occurs. Predictive maintenance is a maintenance procedure in which a maintenance staff measures the operating state of each component device and replaces it with a new component device when a failure sign is found from the measurement result.
Toshio Toyoda, “Progress of Predictive Maintenance Technology and Evolution to Plant Asset Management System PAM (above)”, Plant Engineer, Vol. 33, no. 10, p. 58-61 (2001).

  However, in the maintenance method as described above, a failed component device or a component device in which a failure sign is found is replaced with a new component device after the post maintenance or the predictive maintenance. As a result, immediately after replacement with a new component device, the failure rate of the system increases as in the case of the periodic preventive maintenance described above, based on the characteristics of the initial failure type by the component device. Therefore, there is a problem that the operating rate of the system is lowered, the maintenance cost is increased, and a large amount of damage is caused by the system being stopped.

  The present invention has been made in view of the above circumstances. A system maintenance method, a maintenance device, and a program for realizing a reduction in maintenance costs and a damage amount due to a system stop while improving a system operating rate during system maintenance. The purpose is to provide.

(1) In order to solve the above problem, in a maintenance method for a system having at least one component device including parts, a component device at the start of system operation or a component device that has failed at the time of subsequent maintenance or predictive maintenance Device data input step for inputting model name and usage time, model name, usage time and failure rate of a plurality of component devices are registered, and the plurality of registered configurations based on the model name of the input component device A model extraction step for extracting a component device having the same model name from the devices, a random failure type determination step for determining a component device having a failure rate that has shifted from the failure rate of the extracted component device to a random failure type, and When the usage time of the component device determined by the random failure type determination step is longer than the predetermined boundary time of the component device of the same model name, the determined component device A maintenance method of a system having an interchangeable device selection step of determining that exchangeable construction equipment output.

  Instead of the accidental failure type determination step, there is an initial failure type determination step for determining a component device having an initial failure type failure rate from the failure rate of the extracted component device. When the usage time of the configured component device is longer than the input usage time of the component device, the determined component device may be determined as a replaceable component device.

(2) Further, the system maintenance apparatus according to the present invention is a system maintenance apparatus including at least one component device including parts, and includes a plurality of component device model names, usage times, failure rates, and The storage means for storing the boundary time of the component equipment, the input means for inputting the model name and usage time of the component equipment at the start of system operation, or the component equipment that failed during the post-maintenance and predictive maintenance, and this input Based on the model name of the component device, a model extraction unit that extracts a component device having the same model name from the model names of the plurality of component devices stored in the storage unit, and a random occurrence from the failure rate of the extracted component device Random failure type determination means for determining a component device having a failure rate that has shifted to a failure type, and the usage time of the component device determined by the random failure type determination unit is based on the boundary time of the component device having the same model name And determining operating time determining means for determining long or not, it is a configuration and means for outputting determined long been judged exchangeable construction equipment construction equipment in this determination means.

  Further, instead of the accidental failure type determination means, provided is an initial failure type determination means for determining a component device having an initial failure type failure rate from the extracted failure rate of the component device, and the usage time of the determined component device May be determined as a replaceable component device when it is determined that the component device is longer than the usage time of the component device input from the input unit.

  Note that the processing in the above example can also be realized by a program.

  According to the present invention, a system maintenance method and maintenance that can improve the operating rate of the system by reducing and reducing the maintenance cost and the amount of damage due to the system stoppage by selecting and exchanging the most suitable components during system maintenance. Devices and programs can be provided.

  Hereinafter, prior to describing a maintenance method and a maintenance device of a system according to the present invention, an example of a system to which the maintenance method and the maintenance device are applied will be described with reference to FIG.

FIG. 1 is a configuration diagram illustrating an example of an information processing system that processes a company's core business.
In this information processing system, an intra-company communication network 3 is connected to the Internet 1 via a router 2 having a security function. A system including a plurality of component devices (hereinafter, including parts) is connected to the intra-company communication network 3.

  That is, this information processing system is configured by a plurality of component devices having different processing functions, for example, a Web server group 4, an application server group 5, and a database server group 6. Each component device of each of the server groups 4 to 6 incorporates hardware (components) and software (components) for executing required processing.

  The Web server group 4 is a set of a plurality of Web servers 4a, 4b,... For providing content to the user. The software components of each of the Web servers 4a, 4b,... Include a user interface for determining an agreement for data exchange, Web software for creating and displaying a browsing screen, Web server software for performing content transfer control for user access, and an OS. It includes middleware that provides flexibility with applications and an operating system that provides a user operating environment.

  The application server group 5 is a set of application servers 5a, 5b,... That receives a request from a user and executes predetermined basic business processing using data in a database. As software components of each of the application servers 5a, 5b,..., A service interface, application software, application server software, middleware, operating system, and the like are provided.

  The database server group 6 is a set of database servers 6a, 6b,... That manage a database necessary for core business processing. As software components of each of the database servers 6a, 6b,..., A data access interface, database software, middleware, an operating system, and the like are provided.

  Therefore, the system maintenance method and maintenance apparatus according to the present invention is to perform maintenance of the information processing system shown in FIG.

  Next, when describing an embodiment of a system maintenance method according to the present invention, an example of a result of repeating an experiment on the relationship between the usage time and the failure rate of a database server, for example, 6a, which is a component device of an information processing system This will be described with reference to FIGS. In the figure, the horizontal axis represents the usage time of the database server 6a, and the vertical axis represents the failure rate of 1 of 10,000 minutes.

  FIG. 2 shows a state in which the failure rate of the database server 6a has an initial failure type characteristic in which the failure rate decreases with the passage of time from the initial point of time, and thereafter, a transition to an accidental failure type in which the failure rate becomes constant. ing.

  FIG. 3 shows an initial failure type state in which the failure rate of the database server 6a decreases from the initial point in time.

  Then, regarding the maintenance method by the conventional post maintenance and predictive maintenance, the following results were obtained when the relationship between the failure rate of the database server 6a and the influence of the system was examined in detail.

During post maintenance or predictive maintenance, when the database server 6a, which is a failed component device (including a failure sign), is replaced with a new database server, failures often occur even though not much time has passed since the device replacement. It was found to occur. As a result, the operating rate of this information processing system remained at about 99.88% per four years, resulting in a large amount of damage due to system shutdown. By the way, the sum of the maintenance cost and the amount of damage caused by the system outage is about 25 billion yen per 4 years.
Therefore, with regard to system maintenance, the following maintenance methods were accumulated on a trial and error basis.
Normally, it is replaced with a database server at the time of ex-post maintenance or predictive maintenance. At this time, considering the relationship between the elapsed time and the failure rate shown in FIG. 2, the initial failure in which the failure rate decreases with the passage of time. When the model was replaced with a normally operating database server that was used for about 7500 hours, which was the usage time at the time of transition to the accidental failure type with a constant failure rate, the following results were obtained. System availability increased to approximately 99.90% over 4 years. When the maintenance cost and the amount of damage due to the system stop were added up, it was possible to reduce the cost by about 2.5 billion yen compared to the conventional maintenance method.

  Therefore, in the present invention, the system maintenance method shown in FIG. 4 is carried out in consideration of the relationship between the passage of time and the failure rate shown in FIGS.

(1) Register the model name, usage time (use time) and failure rate of many component devices used in advance, and model name and usage time of the failed component device including a failure sign during subsequent maintenance or predictive maintenance (Failed device data input step S1).

(2) Based on the acquired model names of the component devices, for example, component devices having the same model name are extracted from the model names of a large number of replaceable component devices stored in a database (model extraction step S2).

(3) The failure type is determined from the failure rate associated with the model name of the extracted component device. That is, whether the failure rate of the extracted component device is a component device having a failure rate that changes from an initial failure type in which the failure rate decreases with time to a random failure type in which the failure rate is constant (see FIG. 2) or failure It is determined whether the device has a failure rate of the initial failure type whose rate decreases with time (see FIG. 3) (failure type determination step S3).

(4) Subsequently, when it is determined in the failure type determination step S3 that the component device has shifted to the accidental failure type, the component device has a longer usage time than the boundary time corresponding to about 700 hours described above. For example, the long component device is selected as a replaceable component device and replaced with a failed component device.

  On the other hand, when the failure type determination step S3 determines that the component device has an initial failure type, if the component device has a usage time longer than the usage time of the failed configuration device, the component device having the long usage time is selected. Select and replace the failed component device (replacement component device selection step S4).

  Therefore, according to the maintenance method of the system as described above, the configuration of the same model name that has already been used for a longer time when replacing a failed component device that includes a failure sign during subsequent maintenance or predictive maintenance. By exchanging with the device, the failure rate can be reduced even immediately after the replacement, the system operating rate can be improved, and the cost can be greatly reduced.

  Next, a maintenance apparatus for a system according to the present invention for carrying out the maintenance method will be described with reference to the drawings.

FIG. 5 is a block diagram showing an embodiment of a system maintenance apparatus according to the present invention.
The maintenance device of the system includes a failure device data input unit 11 for inputting a failure sign, a model name of a failed component device, and a usage time at the time of subsequent maintenance and predictive maintenance, a program memory 12, a database 13, and a CPU, for example. The replacement device selection processing unit 14, the data buffer memory 15, and the data output unit 16 are configured.

  The faulty equipment data input unit 11 normally uses an input device 11a such as a keyboard or a mouse, and inputs a fault sign, a model name of the constituent equipment that has failed, and a usage time at the time of subsequent maintenance or predictive maintenance. The maintenance staff also includes means for transmitting and inputting the model name and usage time of the component equipment and receiving the processing result from the site using the communication network 11b such as the corporate network or the Internet.

  The program memory 12 stores a replacement processing program that defines a series of processing procedures for selecting a replacement device. The exchange device processing unit 14 can also be configured with logical circuit elements without using a CPU. In this case, it is not always necessary to use a CPU, and no program memory is required.

  In the database 13, a device management table 13a for registering at least the model name, usage time, and failure rate of the replaceable component device that has already been used in advance, and a boundary time for setting a boundary time for shifting from the initial failure type to the accidental failure type A setting unit 13b is provided. It should be noted that the number of component devices having the same model name is not necessarily limited to one, but naturally includes registration of a plurality of component devices having the same model name, each having a different usage time. At this time, usage time and failure rate data are stored correspondingly for each component device having the same model name.

  When the replacement device selection processing unit 14 is executed by using the replacement processing program, the replacement device selection processing unit 14 functionally includes a model extraction unit 14A, a failure type determination unit 14B, a usage time determination unit 14C, and a data output control unit 14D. Is provided.

  The model extracting unit 14A has a function of extracting a component device having the same model name from the device management table 13a of the database 13 based on the model name and usage time of the failed component device input from the failed device data input unit 11. Yes.

  The failure type determination unit 14B determines a failure type related to the model name of the component device extracted by the model extraction unit 14A.

  There are two types of component failure types. One of them is a component device having a failure rate in which the failure rate of the component device goes through an initial failure type in which the failure rate decreases with the passage of time from the initial time point, and then changes to a random failure type in which the failure rate becomes constant. 2 is a component device. The other one is a component device having an initial failure type failure rate in which the failure rate of the component device decreases with the passage of time from the initial time point, and is a component device belonging to FIG.

  The failure type determination means 14B is a component device having a failure rate shifted from the failure rate associated with the model name of the component device extracted from the device management table 13a of the database 13 to the accidental failure type, or remains in the initial failure type. It is determined whether the component device has a failure rate.

  The usage time determination means 14C is a replaceable configuration if the extracted configuration device is a configuration device having a usage time longer than the usage time of the failed configuration device including a failure sign at the time of post-mortem maintenance or predictive maintenance. It is determined that the device can be used.

  The data output control unit 14D is connected to a display unit, a printer, or a network, and in accordance with an instruction from the failed device data input unit 1, selects any one of the display unit, the printer, and the network, and data related to the component device selected as replaceable Is output.

  The data buffer memory 15 has a function of temporarily storing data input from the failed device data input unit 1 and data being processed.

  The program memory 12, the database 13, and the data buffer memory 15 are provided in an independent state. However, for example, the database 13 may be divided into areas and shared.

  Next, a series of processing examples of the operation of the maintenance device of the system as described above and the replacement processing program will be described with reference to FIG.

  When the maintenance device of the system starts operation, the replacement device selection processing unit 14 reads out the device replacement processing program from the program memory 12 and temporarily stores it in the data buffer memory 15 or the like, and then a series of processing by the replacement processing program. Execute.

  First, the replacement device selection processing unit 14 executes the model extraction unit 14A. This model extraction means 14A relates to the model name and usage time of a component device (hereinafter referred to as a failure component device) that is determined to have a failure sign or a failure at the time of subsequent maintenance or predictive maintenance from the failure device data input unit 11. It is determined whether data has been input (S11). Here, if it is determined that the model name and usage time data of the failed component device have been input, the data is temporarily stored in the data buffer memory 15 and then the device management table 13a of the database 13 based on the model name of the failed component device. In step S12, it is determined whether there is a component device having the same model name among the component devices registered in the device management table 13a.

  When it is determined that there is no component device with the same model name, if there are other component devices with the same model name, that is, if multiple component devices with the same model name are registered, all the same model names The presence / absence of the component equipment is determined (S13).

  If it is determined in step S12 that a component device having the same model name exists, the model name, usage time, and failure rate data of the corresponding component device are extracted from the device management table 13a and stored in the data buffer memory 5. These steps S11 to S13 correspond to a model extraction procedure.

  The replacement device selection processing unit 14 executes the failure type determination unit 14B after extracting the model name, usage time, and failure rate data of the corresponding component device.

  The failure type determination means 14B determines whether the extracted component device having the same model name belongs to the above-described two failure types. That is, the failure type determination means 14B is a component device having a failure rate that has shifted from an initial failure type in which the failure rate of the component device having the extracted model name decreases with time to a random failure type in which the failure rate is constant. It is determined whether or not there is (S14).

  In step S14, when it is determined that the extracted component device is not a component device having a failure rate that has shifted to the accidental failure type, the failure rate of the initial failure type in which the failure rate of the component device decreases with the passage of time continues. It is determined whether or not the component device has a rate (S15). These steps S14 and S15 correspond to a failure type determination procedure.

  If the component device extracted in steps S14 and S15 does not belong to any failure type, the exchange device selection processing unit 14 proceeds to step S13, and whether or not all the component devices having the same model name exist. Judging.

  On the other hand, if it is determined that the replacement device selection processing unit 14 is a component device having a failure rate that has shifted to the accidental failure type in step S14, or the component device having the initial failure type failure rate in step S15. When it is determined that there is, the usage time determination unit 14C is executed.

  When the usage time determination unit 14C determines that the failure type determination unit 14B is a component device having a failure rate shifted to the accidental failure type, the usage time of the component device is the same as that set in the boundary time setting unit 13b. It is determined whether or not the device has a usage time longer than the boundary time associated with the device having the model name (S16). On the other hand, when it is determined that the component device has an initial failure type failure rate, the determined usage time of the component device is compared with the usage time of the failed component device input from the failed device data input unit 1. Then, it is determined whether or not the use time of the determined component device is a component device having a use time longer than the use time of the failed component device (S17). If NO in both steps S16 and S17, the process proceeds to step S13 to determine whether there is another component device having the same model name. These steps S16 and S17 correspond to a usage time determination procedure.

  The exchange device selection processing unit 14 executes the data output control unit 14D when it is determined in step S16 or step S17 that the use time of the replaceable component device is long.

  The data output control unit 14D selects and outputs the data of the replaceable component device whose use time is assumed to be long according to the output instruction input from the failed device data input unit 11, or the network. The request is transmitted through 1b and returned (S18: data output procedure).

  Therefore, the maintenance staff can determine whether a failure or a sign of failure occurs after the ex-post maintenance or predictive maintenance, based on the data of the replaceable component device output by the data output control means 14D, A replaceable component device is obtained from the device holding location, and the failed component device is replaced with a replaceable component device.

  In step S13, when there is no component device having the same model name as the model name of the failed component device in the device management table 13a, data indicating that there is no replaceable component device is output (S19).

  Therefore, according to the embodiment as described above, when a failure predictor or a failed component device is discovered during the post-maintenance or predictive maintenance, it is not a completely new component device but a component device already in use. When a component device having the same model name is extracted and it is determined that the usage time of the extracted component device is longer than a predetermined usage time according to the failure type, the extracted component device is Select and replace the component as a replaceable component. In this way, when a failure sign or a failed component device is replaced with a selected component device during post-mortem maintenance or predictive maintenance, the failure rate of the replaced component device is small, so the system operation rate is improved. Can do. As a result, since the component devices that have already been used are used, maintenance costs can be suppressed, and the influence of the system stop can also be suppressed. Therefore, the total cost can be greatly reduced as compared with the conventional system maintenance.

(Other embodiments)
(1) In the above embodiment, when a failure predictor or a failed component device is found during post-mortem maintenance or predictive maintenance, an optimal component device is selected from the already used component devices and replaced. It is also possible to start the operation of the system by using the optimum component device among the component devices already used as the component device at the time of starting the system operation, not at the time of the ex-post maintenance or predictive maintenance.

(2) In the above embodiment, the application example to the information processing system has been described. However, the system to be applied is not particularly limited. For example, the present invention can be similarly applied to various plant systems such as steel control systems, oil refining systems, and power system control systems, and air traffic control systems.

  In addition, the said embodiment only illustrated one specific example, and does not limit this invention. That is, it goes without saying that various modifications can be made according to the specific application target system without departing from the scope of the invention. In addition, the operations and effects in the above-described embodiment list the most preferable operations and effects resulting from the embodiment, and are not necessarily limited to the description.

The block diagram of the information processing system for demonstrating the example of 1 application object with respect to the maintenance method of the system which concerns on this invention, and a maintenance apparatus. The figure showing the state in which the failure rate of a system component apparatus is shifting from the initial failure type to the accidental failure type. The figure showing that the failure rate of a system component apparatus exists in the state of an initial stage failure type. The figure explaining one Embodiment of the maintenance method of the system which concerns on this invention. The block diagram which shows one Embodiment of the maintenance apparatus of the system which concerns on this invention. The figure explaining operation | movement of the maintenance apparatus of the system which concerns on this invention, and the flow of the process by the program which concerns on this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Internet, 3 ... In-company communication network, 4 ... Web server group, 5 ... Application server group, 6 ... Database server group, 11 ... Failure apparatus data input part, 12 ... Program memory, 13 ... Database, 14 ... Exchange equipment Selection processing unit, 14A ... model extraction unit, 14B ... failure type determination unit, 14C ... usage time determination unit, 14D ... data output control unit, 16 ... data output unit.

Claims (5)

In a maintenance method for a system including at least one component device including parts,
A device data input step for inputting the model name and usage time of a component device at the start of system operation or a component device that failed during post-mortem maintenance or predictive maintenance;
A model in which model names, usage times, and failure rates of a plurality of component devices are registered, and a component device having the same model name is extracted from the plurality of registered component devices based on the input model device name An extraction step;
Random failure type determination step for determining a component device having a failure rate shifted from the failure rate of the extracted component device to a random failure type,
When the usage time of the component device determined by the random failure type determination step is longer than the predetermined boundary time of the component device having the same model name, the determined component device is determined to be a replaceable component device and output. A system maintenance method comprising: a replacement device selection step. A device data input step for inputting the model name and usage time of a component device at the start of system operation or a component device that failed during post-mortem maintenance or predictive maintenance;
A model in which model names, usage times, and failure rates of a plurality of component devices are registered, and a component device having the same model name is extracted from the plurality of registered component devices based on the input model device name An extraction step;
An initial failure type determination step for determining a component device having an initial failure type failure rate from the failure rate of the extracted component devices;
When the usage time of the component device determined in this determination step is longer than the input component device use time, the replacement device selection step of determining and outputting the determined component device as a replaceable component device, and A system maintenance method comprising: In a maintenance apparatus for a system having at least one component including parts,
Storage means for storing in advance model names, usage times, failure rates, and boundary times of each component device;
An input means for inputting the model name and usage time of a component device at the start of system operation or a component device that failed during post-mortem maintenance or predictive maintenance;
Based on the input model name of the component device, model extraction means for extracting the component device having the same model name from the model names of the plurality of component devices stored in the storage unit;
Random failure type determination means for determining a component device having a failure rate shifted from the failure rate of the extracted component device to a random failure type,
Usage time determination means for determining whether the usage time of the component device determined by the random failure type determination means is longer than the boundary time of the component device of the same model name;
A system maintenance apparatus comprising: means for determining and outputting a component device that is determined to be long by the determination unit as a replaceable component device. In a maintenance apparatus for a system having at least one component including parts,
Storage means for storing the model name, usage time, failure rate, and boundary time of each component device of a plurality of component devices;
An input means for inputting the model name and usage time of a component device at the start of system operation or a component device that failed during post-mortem maintenance or predictive maintenance;
Based on the input model name of the component device, model extraction means for extracting the component device having the same model name from the model names of the plurality of component devices stored in the storage unit;
An initial failure type determination means for determining a component device having an initial failure type failure rate from the extracted failure rate of the component device;
Use time determination means for determining whether or not the use time of the component device determined by the initial failure type determination means is longer than the use time of the component device input from the input means;
A system maintenance apparatus comprising: means for determining and outputting a component device that is determined to be long by the determination unit as a replaceable component device. For equipment replacement processing in which the model name, usage time, failure rate, and boundary time of each constituent equipment are registered, and the constituent equipment that can be replaced at the time of system operation or after maintenance or predictive maintenance is selected. A program,
In the computer,
Based on the model name and usage time of the component device at the start of system operation input from the outside, or the component device that failed during the post-maintenance or predictive maintenance, the same among the model names of the plurality of registered component devices The model extraction procedure for extracting the component device of the model name and whether the component device has a failure rate that has shifted from the failure rate of the extracted component device to the accidental failure type or the failure rate of the initial failure type When it is determined that the component has a failure type determination procedure and a failure rate that has shifted to a random failure type by this determination procedure, the usage time of the component device having the failure rate is the initial failure type and the accidental failure. When it is determined by the failure type determination procedure that the component device has a failure rate of the initial failure type by determining whether the time is longer than the boundary time with the mold, the failure rate Configuration machine with A component device that has been determined to be longer by either the second use time determination procedure for determining whether or not the use time is longer than the use time of the component device input from the outside, and both the use time determination procedures A device replacement processing program for executing a data output procedure for outputting as a replaceable component device.

Images