CN107402873B

CN107402873B - MMBF fault statistical system and method

Info

Publication number: CN107402873B
Application number: CN201710627751.XA
Authority: CN
Inventors: 孙晶哲
Original assignee: Shanghai Zpmc Electric Co ltd; Shanghai Zhenghua Heavy Industries Co Ltd
Current assignee: Shanghai Zpmc Electric Co ltd; Shanghai Zhenghua Heavy Industries Co Ltd
Priority date: 2017-07-28
Filing date: 2017-07-28
Publication date: 2020-10-27
Anticipated expiration: 2037-07-28
Also published as: CN107402873A

Abstract

The invention discloses an MMBF fault statistical system and method. The historical database comprises fault points and other related points, and the fault judgment module comprises a TimeCount calculator, a keepTime calculator and a ChangeTimeCount calculator. The fault judgment module periodically reads fault points and other related points from the historical database, and utilizes any one or more of the TimeCount arithmetic unit, the KeepTime arithmetic unit and the ChangeTimeCount arithmetic unit to judge faults so as to generate a plurality of suspected MMBF. The screening module screens one suspected MMBF from the plurality of suspected MMBF to serve as the MMBF within the time period. The fault data required by the invention directly comes from the historical database, thereby saving an intermediate conversion link.

Description

MMBF fault statistical system and method

Technical Field

The present invention relates to a fault statistics system and method, and more particularly, to an MMBF fault statistics system and method.

Background

Most of the existing fault statistical systems are based on a relational database, and few statistical systems are based on a special historical/real-time database. When creating a relational database, which can define a range of possible values for a column of data and further constraints that may be applied to that data value, the SQL language is a standard user and application interface to the relational database. The advantage of a relational database is that it is easily scalable and that after initial database creation, a new data class can be added without the need to modify all existing application software.

However, although the relational database-based data can be read in a consistent manner, most of the data collected in the industry is stored in the historical/real-time database, and if a statistical tool based on the relational database is used, conversion from the historical/real-time data to the relational database is required, which wastes time and increases the uncertainty of the system.

Disclosure of Invention

In view of the above problems in the prior art, an object of the present invention is to provide an MMBF fault statistics system and method.

In order to achieve the purpose, the invention adopts the following technical scheme:

an MMBF fault statistical system comprises a historical database, a fault judgment module and a screening module. The historical database comprises fault points and other related points, and the fault judgment module comprises a TimeCount calculator, a keepTime calculator and a ChangeTimeCount calculator. The fault judgment module periodically reads fault points and other related points from the historical database, and utilizes any one or more of the TimeCount arithmetic unit, the KeepTime arithmetic unit and the ChangeTimeCount arithmetic unit to judge faults so as to generate a plurality of suspected MMBF. The screening module screens one suspected MMBF from the plurality of suspected MMBF to serve as the MMBF within the time period.

Further, the TimeCount operator includes: fault setting time: initially 0; reporting the state by MMBF: initially "unreported"; the computing interface logic of the TimeCount computing device is as follows: 2.1: setting the Fault for more than 0 to 2, otherwise, setting the Fault for 2.6; 2.2: if the MMBF reporting state is 'not reported', continuing, otherwise, turning to 2.4; 2.3: calculating the reset operation times from the Fault setting time to the current calculating time, reporting a suspected MMBF Fault if the times is more than the set times, and setting the MMBF reporting state as 'reporting'; 2.4: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit'; 2.5: exit logic; 2.6: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter; 2.7: and exiting the logic.

Further, the keepdime operator includes: fault setting time: initially 0; reporting the state by MMBF: initially "unreported"; first RESET time: initially 0; the operation interface logic of the KeepTime operator is as follows: 3.1: setting the Fault for more than 0 to 2, otherwise, setting the Fault for 3.8; 3.2: if the MMBF reporting state is 'reporting', exiting the logic, otherwise continuing; 3.3: if the first RESET time is more than 0 to 4, otherwise, 3.5 is reached; 3.4: if the price from the time parameter to the first RESET time is larger than the set value, reporting a suspected MMBF fault, and setting the MMBF reporting state as 'reporting'. Go to 3.6; 3.5: searching for the first-time RESET time from the Fault setting time to the current calculation time (time parameter), wherein if the first-time RESET time is found, the setting is carried out, and if the first-time RESET time is not found, the first-time RESET time is 0; 3.6: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit'; 3.7: exit logic; 3.8: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter; 3.9: and exiting the logic.

Further, the ChangeTimeCount operator includes: last "value" parameter: initially resetting; set of times previously changed from reset to set state: an empty set is initialized; the logic of the computing interface of the ChangeTimeCount operator is as follows: 4.1: if the value parameter is the same as the last value parameter, exiting the logic, otherwise continuing; 4.2: removing records exceeding a set time in the time set; 4.3: if the value parameter is reset, the logic is exited, otherwise, the logic continues; 4.4: adding "time" to the time set; 4.5: counting the number of records in the time set, if the number of records is less than a set value, exiting the logic, otherwise, continuing; 4.6: reporting a suspected MMBF fault; 4.7: and exiting the logic.

Furthermore, the screening module takes out the longest time as statistical time, advances the MMBF time appearing first in a time interval backward for statistical time and then for a minute to obtain statistical deadline, and starts to count the MMBF of the time interval when the operation reaches the statistical deadline.

Further, the screening module resets all points for a set length of time, at which point screening for suspected MMBF is initiated prior to that point.

Further, the fault point is a boolean point, and other relevant points include a reset point and a test point.

In order to achieve the purpose, the invention also adopts the following technical scheme:

an MMBF fault statistical method comprises the following steps: periodically reading fault points and other related points from a historical database, and performing fault judgment by using any one or more of a TimeCount method, a KeepTime method and a ChangeTimeCount method to generate a plurality of suspected MMBF; and screening one suspected MMBF from the plurality of suspected MMBF as the MMBF in the time period.

Further, the TimeCount method includes: setting Fault setting time: initially 0; setting an MMBF reporting state: initially "unreported"; the TimeCount logic is as follows: 9.1: setting the Fault for more than 0 to 2, otherwise, 9.6; 9.2: if the MMBF reporting state is 'not reported', continuing, otherwise, turning to 9.4; 9.3: calculating the reset operation times from the Fault setting time to the current calculating time, reporting a suspected MMBF Fault if the times is more than the set times, and setting the MMBF reporting state as 'reporting'; 9.4: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit'; 9.5: exit logic; 9.6: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter; 9.7: and exiting the logic.

Further, the keepdime method includes: setting Fault setting time: initially 0; setting an MMBF reporting state: initially "unreported"; setting a first RESET time: initially 0; the keepdime operation logic is as follows: 10.1: setting the Fault for more than 0 to 2, otherwise, setting the Fault for 10.8; 10.2: if the MMBF reporting state is 'reporting', exiting the logic, otherwise continuing; 10.3: if the first RESET time is more than 0 to 4, otherwise, 10.5 is reached; 10.4: if the price from the time parameter to the first RESET time is larger than the set value, reporting a suspected MMBF fault, and setting the MMBF reporting state as 'reporting'. Go to 10.6; 10.5: searching for a first RESET time from the Fault setting time to the current calculation time, wherein if the RESET time can be found, the RESET time is set, and if the RESET time can not be found, the RESET time for the first time is 0; 10.6: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit'; 10.7: exit logic; 10.8: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter; 10.9: and exiting the logic.

Further, the ChangeTimeCount method includes: setting the parameter of 'value' at the last time: initially resetting; set the time to change from reset to set state before setting: an empty set is initialized; the ChangeTimeCount operation logic is as follows: 11.1: if the value parameter is the same as the last value parameter, exiting the logic, otherwise continuing; 11.2: removing records exceeding a set time in the time set; 11.3: if the value parameter is reset, the logic is exited, otherwise, the logic continues; 11.4: adding "time" to the time set; 11.5: counting the number of records in the time set, if the number of records is less than a set value, exiting the logic, otherwise, continuing; 11.6: reporting a suspected MMBF fault; 11.7: and exiting the logic.

Further, the method for screening the suspected MMBF from the plurality of suspected MMBF comprises the following steps: taking out the longest time as statistical time; the MMBF time appearing first in a time interval is pushed backwards for a statistical time and then for a minute, and a statistical cut-off time is obtained; when the operation reaches the statistical cutoff time, the MMBF of the period is counted.

Further, the method for screening the suspected MMBF from the plurality of suspected MMBF comprises the following steps: all the points are reset and continue for a set length of time, at which point screening for suspected MMBF is initiated prior to this point.

In the technical scheme, the fault data required in the MMBF fault statistical system and method provided by the invention directly comes from the historical database, so that an intermediate conversion link is omitted.

Drawings

FIG. 1 is a system block diagram of the present invention;

fig. 2 is a flow chart of the present invention.

Detailed Description

The technical scheme of the invention is further explained by combining the drawings and the embodiment.

The MMBF of the present invention means: the MMBF fault is a fault needing to be counted in the MMBF value calculation process. The general idea of the invention is to read the original data from the historical database, and compared with the relational database, the historical database has the advantages of large storage capacity, high query speed and the like. When MMBF fault statistics is carried out, a larger amount of data can be read at the same time for analysis.

The invention reads all fault point (fault point, must be Boolean type point) names to be analyzed and other related point names (reset point, test point, etc.) from the engineering file (XML). And then periodically reading relevant data from the historical database for analysis. If the actual fault is analyzed, the information of the fault occurrence time, the fault point name, the fault type and the like is stored in a set relational database for other systems to read.

Referring to fig. 1 and 2, the present invention first discloses an MMBF fault statistical system, which mainly includes a history database 1, a screening module 2, and a fault determination module 3. The historical database 1 further includes a failure point and other related points, and the failure determination module 3 further includes a TimeCount operator 4, a keepdime operator 5, and a ChangeTimeCount operator 6.

The above modules execute the following processes:

s1: the failure determination module 3 periodically reads failure points and other relevant points from the historical database 1,

s2: the failure determination module 3 performs failure determination using any one or more of the TimeCount operator 4, the keepdime operator 5, and the ChangeTimeCount operator 6,

s3: the failure determination module 3 generates a plurality of suspected MMBFs.

S4: the screening module 2 screens one suspected MMBF from the plurality of suspected MMBF as the MMBF in the time period.

The MMBF fault judgment system of the invention is composed of a plurality of layers of a system, a grouping, a data point and an arithmetic unit. The "system" is used as a container of "packets" to implement the function of simultaneous operation of a plurality of packets. The 'grouping' is an independent unit for judging the MMBF fault, and the groups are independent and do not influence each other. The "data points" are added to the "packets" to complete the determination of a suspected fault. The same "data point" may be added to different "groupings". A "data point" is a logical Boolean quantity that can take only two values, set and reset. The 'operator' is responsible for judging whether the fault occurs or not by using the data of the 'data point'. A "data point" can be used for multiple types of fault judgment by using multiple different types of "operators". The "arithmetic unit" corresponds to a failure determination method, and different failure determination methods are determined by different "arithmetic units".

The overall logic of the invention:

the method periodically acquires data from a historical database and a relational database, wherein the period is generally 5 to 10 minutes, the acquired data are transmitted into an MMBF system object, and the system object classifies and caches the data according to different data points. After all the data in the period are transmitted into the system object, the system object batches the data in smaller time slices (10 seconds) and transmits the data to each data point, and meanwhile, the data point transmits the data to each arithmetic unit. If the 'arithmetic unit' generates a suspected fault, the suspected MMBF fault is reported to the group where the arithmetic unit is located. When all data in the week are distributed, the system object traverses and calls the grouping object to acquire the MMBF fault judged by the grouping object. When the 'system' object distributes data to 'data points' in short time slices, if a certain data point has no data change in the time slice, the data of the previous time slice is distributed to the data point, so that the opportunity of operation of each data point in each time slice can be ensured.

'packet' generation MMBF fault logic

First, whether a judgment period of MMBF fault is finished or not is judged. In the execution process of the method, the data points are submitted to a plurality of suspected MMBF faults in groups, but only one MMBF fault is effective in the same period, so whether the period is finished or not is judged firstly.

1: and judging the value of the data point in the grouping to be the set point number. The number of points is greater than 0 to 2, and the number of points is 0 to 3.

2: the point number is larger than 0, and "no failure state start time" is set to 0. The determination period is not over. And exiting the logic.

3: the point number is 0, which indicates that no fault occurs in the calculation period. "no fault state start time" is judged. "no fault condition start time" is 0 to 4; the "no fault condition start time" is greater than 0 to 5.

4: the "no-fault state start time" is set as the end time of this calculation cycle. The determination period is not over. And exiting the logic.

5: the "no fault state duration" is judged. The "no fault condition duration" is greater than the set value (60 seconds) to 6, otherwise to 7.

6: and (5) ending the judgment period of the MMBF fault, and recording the ending state of the judgment period. And exiting the logic.

7: at this time, some system points are set (state points such as the forced calculation period is ended, the control process is ended and the like), and the judgment period of the MMBF fault is ended. Otherwise, the period is judged not to be finished. And exiting the logic.

And then calculating the MMBF fault, and analyzing an MMBF fault from a plurality of suspected MMBF faults submitted by the data points. The conditions need to be satisfied: the time value is the smallest (occurring first) and the time to the end of the calculation cycle is greater than a certain value (the longest determination time of various "operators" is 1 minute again). While different logics are performed according to the recorded faults.

1: if there is a recorded fault of 2, otherwise 4.

2: and judging whether the MMBF fault period is finished or not, and if so, exiting the logic. Otherwise to 3.

3: and reporting the end of an MMBF fault period of the system. This "packet" reset. And exiting the logic.

4: and (4) analyzing an MMBF fault in the suspected MMBF fault, if the MMBF fault can be generated, reporting the MMBF fault of the system, and recording the fault. And if the information cannot be generated, the judgment is quitted.

5: if the judgment period of the MMBF fault is not finished, reporting that one MMBF fault period of the system is finished, and resetting the grouping.

'packet' reset logic

1: clearing the recorded fault;

2: clearing all suspected MMBF faults in the cache;

3: all "data point" states are reset.

Note that: the "data point" will only report at most one suspected MMBF failure to the "packet" before reset. More than one suspected MMBF fault is not reported.

Arithmetic logic of arithmetic unit

All "operator" objects provide an operator interface that is called by the "system" object to transfer data. This interface has two parameters: "time" and "value" (logical type). Both parameters represent the value of the "data point" to which they belong at a certain time.

The main decision logic of the arithmetic unit is realized in the interface. The failure determination method in the present invention is of the following type:

1) TimeCount method: the variables can be set: count (limit number), Time (contract Time). When a fault point is set, the fault point starts to be monitored, the number of reset operations occurring later is recorded, if the number of reset operations occurring reaches the set number in a time less than the set time, and the fault point is still set for a set time (within 5 seconds) after the reset operations of the fault point reach the set number, a suspected MMBF is generated. And reporting the MMBF occurrence time when the fault is changed into a setting state. If in the monitoring state. If the fault point is reset, then the computation logic resets and the fault is recalculated. If the reset operation is not generated for the set times in the monitoring state and the set time, the MMBF will not be reported. After the suspected MMBF is generated, the logic does not calculate the fault point until the fault point is reset, and the logic recalculates the fault point.

The TimeCount operator of the present invention performs the TimeCount mode failure determination, and has the following attributes:

fault setting time: initially 0.

Reporting the state by MMBF: initially "not reported".

The operation interface logic is as follows:

1: the Fault set time is greater than 0 to 2, otherwise to 6.

2: if the MMBF reporting state is 'not reported', continuing, otherwise, turning to 4.

3: and calculating the reset operation times from the Fault setting time to the current calculation time (time parameter), reporting a suspected MMBF Fault if the times are more than the set times, and setting the MMBF reporting state as 'reporting'.

4: and judging the value parameter, and resetting the arithmetic unit if the value parameter is reset.

5: and exiting the logic.

6: judging the 'value' parameter, and if the 'value' parameter is set, recording the setting of Fault as the 'time' parameter.

7: and exiting the logic.

2) ChangeTimeCount mode: the variables can be set: count (limit number), Time (contract Time). At a fault point, setting change of set times (false is changed into true) occurs at a set time, a suspected MMBF is generated once, and the recorded set times are cleared. For example, if the set number is 5 and the set time is 10 minutes, if 6 times of setting occur within 10 minutes, the suspected MMBF will be reported when the setting occurs for the fifth time, then the internal counter will be cleared, and when the setting is detected for the sixth time, the internal counter will be not 6. In addition, the information of each setting is recorded, and only the setting information in the latest 10 minutes (setting time) is reserved, so that the condition that the MMBF is generated without reaching the setting times can be ensured. And when the suspected MMBF is reported, considering the time of the MMBF as the time of the first setting in the set setting times.

The ChangeTimeCount operator of the present invention performs a failure determination in the ChangeTimeCount mode, and has the following attributes: :

last "value" parameter: initially at reset.

Set of times previously changed from reset to set state: initially as an empty set.

The operation interface logic is as follows:

1: if the "value" parameter is the same as the last "value" parameter, the logic exits, otherwise it continues.

2: records exceeding a set time are removed from the time set.

3: if the "value" parameter is reset, the logic exits, otherwise it continues.

4: add "time" to the time set.

5: and counting the number of records in the time set, exiting the logic if the number of records in the time set is less than a set value, and continuing the logic if the number of records in the time set is not less than the set value.

6: reporting a suspected MMBF fault.

7: and exiting the logic.

3) Keep time mode: the variables can be set: time (contracted Time). When a fault point is set, the monitoring of the fault point is started, and if a reset operation occurs later, the timing is started. If the timed time reaches the set time and the fault is still in the set state, a suspected MMBF is generated. After the suspected MMBF is generated, the logic does not calculate the fault until the fault point is reset, and the logic recalculates the fault point. If the fault becomes complex in the monitor or timer state, the computation logic resets and recalculates the fault point. If the fault does not occur a reset operation in the monitor state, then the MMBF will not be generated.

The keepdime operator of the present invention performs failure determination in the keepdime scheme, and has the following attributes:

fault setting time: initially 0.

Reporting the state by MMBF: initially "not reported".

First RESET time: initially 0.

The operation interface logic is as follows:

1: the Fault set time is greater than 0 to 2, otherwise to 8.

2: if the MMBF reporting state is 'reporting', exiting the logic, otherwise, continuing.

3: if the first RESET time is greater than 0 to 4, otherwise to 5.

4: if the price from the time parameter to the first RESET time is larger than the set value, reporting a suspected MMBF fault, and setting the MMBF reporting state as 'reporting'. Go to 6.

5: the first RESET time from the Fault set time to the current calculation time (the "time" parameter) is found, if found, the setting is set, if not found, the first RESET time is 0.

6: and judging the value parameter, and resetting the arithmetic unit if the value parameter is reset.

7: and exiting the logic.

8: judging the 'value' parameter, and if the 'value' parameter is set, recording the setting of Fault as the 'time' parameter.

9: and exiting the logic.

In the present invention, each trigger type can define the conditions of multiple triggers according to actual needs (i.e. different defined times or different appointed time). For a single fault, multiple different types of trigger conditions may be used, but the same type of trigger condition is limited to one. When one point fails, a reset operation is required, and only the failure point which cannot be recovered by the reset operation can be possibly included in the MMBF statistics.

In the invention, each fault point can be set with a plurality of judging modes, and when the data of a certain fault point is judged and the set fault judging characteristic is met, a suspected fault is generated. And after the suspected fault is judged, if the suspected fault is confirmed to be an MMBF fault, reporting, namely: the related information is stored in a relational database. The MMBF fault reporting rule is as follows:

the MMBF reported by the judgment logic is suspected MMBF, and the system stores all the generated suspected MMBF. Only the suspected MMBF at the same time keeps the highest level. If multiple suspected MMBF of the same level are generated at the same time, only one of the MMBF counted first is reserved.

And screening an MMBF which is used in the time period when a certain condition is met, adding the MMBF into the database, and clearing all suspected MMBF stored in the time period.

The conditions for starting the screening of MMBF were as follows:

(1) the conditions set in the system, some of which have set times, take out the longest one as the statistical time. The first MMBF time to occur within a time period is advanced backward by a statistical time plus one minute to obtain a statistical deadline. When the system operation reaches this statistical deadline, the system starts to count the MMBF for this period.

(2) All points in the system are reset for a set length of time at which point the system begins screening for suspected MMBF up to this point.

When the complaint condition is met, the system starts to screen the MMBF, screens one MMBF, and after the MMBF is added into the database, the system can make system reset judgment, and new MMBF can not be generated before the system is not reset. The system reset condition is that all points in the system are reset and the set time length is kept, namely, the condition of starting to screen the MMBF, and when the condition 2 is generated, the system simultaneously reports the MMBF and resets.

When the system starts to screen the MMBF, one piece of logging data is screened from all the reported suspected MMBF, other suspected MMBF are taken as the link of the MMBF, and the screening rule is as follows: and taking the suspected MMBF 5 seconds before all the suspected MMBF, screening the MMBF with the highest grade in 5 seconds, and recording the MMBF into the database.

In summary, the advantages of the invention are as follows:

1, the judgment mode is various, and faults cannot be missed.

And 2, when a plurality of points simultaneously generate faults, the counting is not repeated. Most of the simultaneous multiple faults are from one actual fault, so that multiple reports are not needed.

And 3, designing the level of the fault point, wherein the fault of a high level can cause the fault of a low level. Only one highest level fault is reported each time, which is beneficial to positioning of practical problems.

It should be understood by those skilled in the art that the above embodiments are only for illustrating the present invention and are not to be used as a limitation of the present invention, and that changes and modifications to the above described embodiments are within the scope of the claims of the present invention as long as they are within the spirit and scope of the present invention.

Claims

1. An MMBF fault statistics system, comprising:

the system comprises a historical database, a fault judgment module and a screening module;

the history database comprises fault points and other relevant points, wherein the other relevant points comprise reset points and test points,

the fault judgment module comprises a TimeCount arithmetic unit, a KeepTime arithmetic unit and a ChangeTimeCount arithmetic unit;

the fault judgment module periodically reads fault points and other related points from a historical database, and utilizes any one or more of a TimeCount arithmetic unit, a KeepTime arithmetic unit and a ChangeTimeCount arithmetic unit to judge faults so as to generate a plurality of suspected MMBF;

the screening module screens one suspected MMBF from the plurality of suspected MMBF as the MMBF within the time period,

the screening module takes out the longest time as statistical time, pushes the MMBF time appearing first in a time interval backward for the statistical time and then adds the MMBF time for one minute to obtain statistical cut-off time, and starts to count the MMBF time in the time interval when the calculation reaches the statistical cut-off time;

the TimeCount arithmetic unit judges the fault by utilizing the TimeCount mode, and sets variables as follows: count is the limited times, Time is the appointed Time; when a fault point is set, monitoring the fault point, recording the number of reset operations occurring later, and if the number of reset operations occurring reaches the set number in a time less than the set time, and the fault point is still set for a set time after the reset operations of the fault point reach the set number, generating a suspected MMBF; the reported MMBF occurrence time is the time when the fault is changed into a setting state, if the fault point is reset in a monitoring state, the logic is reset, and the fault point is recalculated; if the reset operation of the set times is not generated in the set time in the monitoring state, the MMBF is not reported; after the suspected MMBF is generated, the logic does not calculate the fault point again until the fault point is reset, the logic recalculates the fault point,

the TimeCount operator performs a TimeCount mode failure determination, and has the following attributes:

fault setting time: initially 0;

reporting the state by MMBF: initially "unreported";

the computing interface logic of the TimeCount computing device is as follows:

2.1: setting the Fault for more than 0 to 2.2, otherwise to 2.6;

2.2: if the MMBF reporting state is 'not reported', continuing, otherwise, turning to 2.4;

2.3: calculating the reset operation times from the Fault setting time to the current calculating time, reporting a suspected MMBF Fault if the times is more than the set times, and setting the MMBF reporting state as 'reporting';

2.4: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit';

2.5: exit logic;

2.6: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter;

2.7: exit logic; and/or

The keepdime operator performs failure determination using keepdime: setting variables: the Time is appointed Time, when a fault point is set, the fault point is monitored, and if reset operation occurs later, timing is started; if the timing time reaches the set time and the fault is still in a set state, generating a suspected MMBF; the reported MMBF generation time is the time when the fault is changed into a setting state; after the suspected MMBF is generated, the logic does not calculate the fault point until the fault point is reset, and the logic recalculates the fault point; if the fault is reset in the monitoring or timing state, the computation logic resets and recalculates the fault point; if the fault does not occur a reset operation in the monitor state, no MMBF will be generated,

the keepdime operator performs keepdime mode failure determination, and has the following attributes:

fault setting time: initially 0;

reporting the state by MMBF: initially "unreported";

first RESET time: initially 0;

the operation interface logic of the KeepTime operator is as follows:

3.1: setting the Fault for more than 0 to 3.2, or else to 3.8;

3.2: if the MMBF reporting state is 'reporting', exiting the logic, otherwise continuing;

3.3: if the first RESET time is more than 0, the RESET time reaches 3.4, otherwise, the RESET time reaches 3.5;

3.4: if the price of the 'time' parameter to the first RESET time is larger than a set value, reporting a suspected MMBF fault, setting the MMBF reporting state as 'reporting', and turning to 3.6;

3.5: searching for the first-time RESET time from the Fault setting time to the current calculation time (time parameter), wherein if the first-time RESET time is found, the setting is carried out, and if the first-time RESET time is not found, the first-time RESET time is 0;

3.6: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit';

3.7: exit logic;

3.8: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter;

3.9: exit logic; and/or

The ChangeTimeCount operator performs failure determination using a ChangeTimeCount method; setting variables: count is the limited times, Time is the appointed Time; setting change of a fault point for a set time is generated at a set time, a suspected MMBF is generated, and the recorded set time is cleared; recording the information of each setting, and only reserving the setting information within the set time, thereby ensuring that the condition that the MMBF is generated without reaching the set times is not generated; when the suspected MMBF is reported, the time of the MMBF is the time of the first setting in the set setting times,

the ChangeTimeCount operator performs a ChangeTimeCount mode failure determination, and has the following attributes:

last "value" parameter: initially resetting;

set of times previously changed from reset to set state: an empty set is initialized;

the logic of the computing interface of the ChangeTimeCount operator is as follows:

4.1: if the value parameter is the same as the last value parameter, exiting the logic, otherwise continuing;

4.2: removing records exceeding a set time in the time set;

4.3: if the value parameter is reset, the logic is exited, otherwise, the logic continues;

4.4: adding "time" to the time set;

4.5: counting the number of records in the time set, if the number of records is less than a set value, exiting the logic, otherwise, continuing;

4.6: reporting a suspected MMBF fault;

4.7: and exiting the logic.

2. The MMBF fault statistics system of claim 1, wherein the screening module resets all points for a set length of time at which screening for suspected MMBF at a time prior to the time begins.

3. The MMBF fault statistics system according to claim 1, where the fault points are boolean type points.

4. An MMBF fault statistical method is characterized by comprising the following steps:

periodically reading fault points and other related points from a historical database, and performing fault judgment by using any one or more of a TimeCount method, a KeepTime method and a ChangeTimeCount method to generate a plurality of suspected MMBF;

screening one of the suspected MMBF as the MMBF in the time period,

the method for screening the MMBF from the plurality of suspected MMBF comprises the following steps:

taking out the longest time as statistical time;

the MMBF time appearing first in a time interval is pushed backwards for a statistical time and then for a minute, and a statistical cut-off time is obtained;

when the operation reaches the statistical cut-off time, beginning to count the MMBF of the time interval;

the other relevant points comprise a reset point and a test point; the TimeCount method performs a failure determination of the TimeCount mode, and has the following attributes:

setting Fault setting time: initially 0;

setting an MMBF reporting state: initially "unreported";

the TimeCount logic is as follows:

9.1: setting the Fault for more than 0 to 9.2, otherwise to 9.6;

9.2: if the MMBF reporting state is 'not reported', continuing, otherwise, turning to 9.4;

9.3: calculating the reset operation times from the Fault setting time to the current calculating time, reporting a suspected MMBF Fault if the times is more than the set times, and setting the MMBF reporting state as 'reporting';

9.4: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit';

9.5: exit logic;

9.6: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter;

9.7: exit logic; the keepdime method performs failure determination in the keepdime mode, and has the following attributes:

setting Fault setting time: initially 0;

setting an MMBF reporting state: initially "unreported";

setting a first RESET time: initially 0;

the keepdime operation logic is as follows:

10.1: setting the Fault for more than 0 to 10.2, otherwise to 10.8;

10.2: if the MMBF reporting state is 'reporting', exiting the logic, otherwise continuing;

10.3: if the first RESET time is more than 0, 10.4 is reached, otherwise 10.5 is reached;

10.4: if the price of the 'time' parameter to the first RESET time is larger than a set value, reporting a suspected MMBF fault, setting the MMBF reporting state as 'reporting', and turning to 10.6;

10.5: searching for a first RESET time from the Fault setting time to the current calculation time, wherein if the RESET time can be found, the RESET time is set, and if the RESET time can not be found, the RESET time for the first time is 0;

10.6: judging the 'value' parameter, if the 'value' parameter is reset, resetting the 'arithmetic unit';

10.7: exit logic;

10.8: judging a value parameter, and if the value parameter is set, recording the Fault set as a time parameter;

10.9: exit logic; the ChangeTimeCount method performs failure determination in the ChangeTimeCount mode, and has the following attributes:

setting the parameter of 'value' at the last time: initially resetting;

set the time to change from reset to set state before setting: an empty set is initialized;

the ChangeTimeCount operation logic is as follows:

11.1: if the value parameter is the same as the last value parameter, exiting the logic, otherwise continuing;

11.2: removing records exceeding a set time in the time set;

11.3: if the value parameter is reset, the logic is exited, otherwise, the logic continues;

11.4: adding "time" to the time set;

11.5: counting the number of records in the time set, if the number of records is less than a set value, exiting the logic, otherwise, continuing;

11.6: reporting a suspected MMBF fault;

11.7: and exiting the logic.

5. The MMBF fault statistics method of claim 4, wherein the fault points are Boolean type points.

6. The MMBF fault statistics method of claim 4, wherein the method of screening the suspected MMBF for the time period comprises:

all the points are reset and continue for a set length of time, at which point screening for suspected MMBF is initiated prior to this point.