Disclosure of Invention
In view of the above, the present invention provides a method for overhauling power plant equipment, so as to solve the problems that the existing overhauling strategy is not highly correlated with the fault and the overhauling task, and the overhauling strategy is disconnected from the reliability analysis. A second object of the invention is to provide a system for servicing power plant equipment.
In order to achieve the first object, the invention provides the following technical scheme:
a method of servicing a power plant device, the method comprising:
acquiring maintenance record data of power plant equipment in real time according to a preset period, wherein the maintenance record data comprises equipment, components, fault modes, maintenance duration and fault time;
processing the data of the maintenance record to obtain a distribution function F of a failure moden(X), wherein n is the number of failure modes, and X is a vector consisting of the equipment, the component, the overhaul mode, the overhaul duration, and the failure time;
processing the data of the maintenance record to obtain a distribution function f of reliabilityn(t), wherein n is the number of failure modes and t is time;
respectively carrying out distribution function F on the fault modes according to the obtained influence coefficients of the running environment and the state of the equipment on the reliability of the fault modesn(X) and a distribution function f of said reliabilitiesn(t) correcting to obtain the distribution function F of the corrected failure modesn' (X) and distribution function f of corrected reliabilityn'(t);
Obtaining L a loss function generated after an equipment failure mode occursnWherein n is the number of failure modes;
according to the loss function L
nAnd a distribution function F of said corrected failure modes
n' (X) obtaining a first function of economic loss probability distribution
According to the loss function L
nAnd a distribution function f of said corrected reliabilities
n' (t) obtaining a second function of economic loss probability distribution
Seeking L a first function corresponding to the economic loss probability distribution by a loop iteration methodMRelative first minimum equipment failure loss value CM;
According to the input maintenance strategy data, carrying out a second function l on the economic loss probability distributionRThe data processing is carried out to obtain an equipment fault loss distribution function LRAnd seeking a second minimum equipment fault loss value C by a loop iteration methodR;
Calculating r-C according to the same maintenance strategyR/CMStopping iteration when r is 0.9-1.1, and setting the final maintenance strategy to be the current first minimum equipment fault loss value CMAnd said second minimum equipment failure loss value CRAnd any corresponding maintenance strategy is utilized to carry out maintenance on the power plant equipment.
Preferably, the second function l for the economic loss probability distribution is based on the inputted overhaul strategy dataRThe data processing is carried out to obtain an equipment fault loss distribution function LRAnd seeking a second minimum equipment fault loss value C by a loop iteration methodRPreviously, the method further comprises:
obtaining the future planned generating capacity of the unit
Wherein p (t) is a function of the power and time of the unit, t1For optimum start time of overhaul, t2For the maintenance end time, the maintenance time length T is T2-t1;
Best start time t for overhaul
1For planning the generation of electricity in the future of the unit
The time corresponding to the minimum;
the overhaul strategy data comprises the equipment, the components, the overhaul mode and the overhaul duration T.
Preferably, said second function/for said economic loss probability distributionRThe data processing is carried out to obtain an equipment fault loss distribution function LRThe method specifically comprises the following steps:
according to the second function l of the economic loss probability distributionRAnd obtaining an equipment fault loss distribution function L by the overhaul strategy data through a Monte Carlo simulation methodR。
Preferably, the data processing is performed on the overhaul record data to obtain a distribution function F of the failure moden(X), specifically including:
classifying historical samples according to six attributes of the equipment, the component, the fault mode, the overhaul duration and the fault time;
taking the distribution rule of the fault mode as a problem to be solved, and respectively taking the five attributes of the equipment, the component, the overhaul mode, the overhaul duration and the fault time as root nodes to construct a decision tree of the distribution rule of the fault mode;
respectively calculating the information gains of the five attributes of the equipment, the component, the overhaul mode, the overhaul duration and the fault time to obtain a distribution function F of a fault moden(X)。
Preferably, the overhaul record data is subjected to data processing to obtain a distribution function f of reliabilityn(t), specifically including:
carrying out data fitting on the maintenance record data by Weibull distribution/normal distribution/Gaussian distribution to obtainDistribution function f to reliabilityn(t)。
The invention provides a maintenance system of power plant equipment, which comprises:
the maintenance record data acquisition module is used for acquiring maintenance record data of the power plant equipment in real time according to a preset period, wherein the maintenance record data comprises equipment, components, a fault mode, a maintenance mode, maintenance time and fault time;
a fault mode distribution function processing module for processing the data of the maintenance record to obtain a fault mode distribution function Fn(X), wherein n is the number of failure modes, and X is a vector consisting of the equipment, the component, the overhaul mode, the overhaul duration, and the failure time;
a reliability distribution function processing module for processing the maintenance record data to obtain a reliability distribution function fn(t); wherein n is the number of the failure modes, and t is time;
an influence coefficient correction module for respectively performing distribution function F on the fault mode according to the obtained influence coefficients of the equipment operating environment and the state on the reliability of the fault moden(X) and a distribution function f of said reliabilitiesn(t) correcting to obtain the distribution function F of the corrected failure modesn' (X) and distribution function f of corrected reliabilityn'(t);
A loss function obtaining module for obtaining a loss function L generated after the device failure mode occursnWherein n is the number of failure modes;
an economic loss probability distribution first function processing module for processing the economic loss probability distribution according to the loss function L
nAnd a distribution function F of said corrected failure modes
n' (X) obtaining a first function of economic loss probability distribution
A second function processing module for economic loss probability distribution according to the loss function L
nAnd a distribution function f of said corrected reliabilities
n' (t) obtaining a second function of economic loss probability distribution
A first minimum equipment failure loss value calculation module for searching a first function L of the economic loss probability distribution by a loop iteration methodMRelative first minimum equipment failure loss value CM;
A second minimum equipment fault loss value calculation module for calculating a second function l of the economic loss probability distribution according to the input maintenance strategy dataRThe data processing is carried out to obtain an equipment fault loss distribution function LRAnd seeking a second minimum equipment fault loss value C by a loop iteration methodR;
A ratio calculation module for calculating r ═ C according to the same maintenance strategyR/CMStopping iteration when r is 0.9-1.1, and setting the final maintenance strategy to be the current first minimum equipment fault loss value CMAnd said second minimum equipment failure loss value CRAnd any corresponding maintenance strategy is utilized to carry out maintenance on the power plant equipment.
Preferably, the system further comprises:
the maintenance time calculation module is used for acquiring the future planned generated energy of the unit
Wherein p (t) is a function of the power and time of the unit, t
1For optimum start time of overhaul, t
2For the maintenance end time, the maintenance time length T is T
2-t
1(ii) a Best start time t for overhaul
1For planning the generation of electricity in the future of the unit
The time corresponding to the minimum; the overhaul strategy data comprises the equipment, the components, the overhaul mode and the overhaul duration T.
Preferably, the second minimum device fault loss value calculation module is specifically configured to:
according to the second function l of the economic loss probability distributionRAnd obtaining an equipment fault loss distribution function L by the overhaul strategy data through a Monte Carlo simulation methodR。
Preferably, the failure mode distribution function processing module is specifically configured to:
classifying historical samples according to six attributes of the equipment, the component, the fault mode, the overhaul duration and the fault time;
taking the distribution rule of the fault mode as a problem to be solved, and respectively taking the five attributes of the equipment, the component, the overhaul mode, the overhaul duration and the fault time as root nodes to construct a decision tree of the distribution rule of the fault mode;
respectively calculating the information gains of the five attributes of the equipment, the component, the overhaul mode, the overhaul duration and the fault time to obtain a distribution function F of a fault moden(X)。
The present invention also provides an apparatus comprising a memory and a processor, wherein:
the memory is used for storing a computer program;
the processor is configured to execute the computer program to implement the method for overhauling the power plant equipment according to any of the above embodiments.
The method for overhauling the power plant equipment comprises the steps of acquiring overhauling record data of the power plant equipment in real time according to a preset period, wherein the overhauling record data comprises equipment, components, a fault mode, an overhauling mode, overhauling time and fault time; processing the data of the maintenance record to obtain a distribution function F of the failure mode
n(X), wherein n is the number of failure modes, and X is a vector consisting of equipment, components, a maintenance mode, maintenance time and failure time; processing the data of the maintenance record to obtain a distribution function f of reliability
n(t), wherein n is the number of failure modesT is time; respectively carrying out distribution function F on the fault modes according to the obtained influence coefficients of the running environment and the state of the equipment on the reliability of the fault modes
n(X) and distribution function f of reliability
n(t) correcting to obtain the distribution function F of the corrected failure modes
n' (X) and distribution function f of corrected reliability
n' (t) obtaining a loss function L generated after the occurrence of the equipment failure mode
nWhere n is the number of failure modes, according to a loss function L
nAnd distribution function F of corrected failure modes
n' (X) obtaining a first function of economic loss probability distribution
According to a loss function L
nAnd distribution function f of corrected reliability
n' (t) obtaining a second function of economic loss probability distribution
Seeking and economic loss probability distribution first function L through loop iteration method
MRelative first minimum equipment failure loss value C
M(ii) a According to the input maintenance strategy data, a second function l is given to the probability distribution of economic loss
RThe data processing is carried out to obtain an equipment fault loss distribution function L
RAnd seeking a second minimum equipment fault loss value C by a loop iteration method
R(ii) a Calculating r-C according to the same maintenance strategy
R/C
MAnd stopping iteration when r is 0.9-1.1, and finally, setting the maintenance strategy as the current first minimum equipment fault loss value C
MAnd a second minimum equipment failure loss value C
RAnd (4) any corresponding maintenance strategy is used for maintaining the power plant equipment by utilizing the final maintenance strategy.
By applying the method and the system for overhauling the power plant equipment, provided by the invention, the data of the overhauling record data is processed to obtain the distribution function of the failure mode, and the loss function generated after the failure mode occurs is combined with the distribution function of the failure mode to form a first function of the economic loss probability distribution to obtain a first minimum equipment failure loss value; the data of the maintenance record data are processed to obtain a distribution function of reliability, and a loss function generated after the fault mode occurs is combined with the distribution function of reliability to form a second function of economic loss probability distribution to obtain a second minimum equipment fault loss value; and stopping iteration by enabling the first minimum equipment fault loss value and the second minimum equipment fault loss value to meet a certain ratio, wherein the final maintenance strategy is any one detection strategy corresponding to the current first minimum equipment fault loss value and the current second minimum equipment fault loss value, and the final maintenance strategy is used for maintaining the power plant equipment, so that the accuracy is improved.
The method combines the maintenance strategy, the fault and the maintenance task, combines the reliability distribution function of the equipment fault mode with the economic loss when the fault mode occurs, obtains the economic loss probability distribution function of the equipment, converts the reliability distribution of the equipment into the economic loss distribution of the equipment fault, and meets the requirements of users. The invention also provides a device.
Detailed Description
The embodiment of the invention discloses a maintenance method of power plant equipment, which aims to solve the problems that the existing maintenance strategy is not high in relevance with faults and maintenance tasks, and the maintenance strategy is disconnected with reliability analysis.
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a schematic flow chart of a method for overhauling power plant equipment according to an embodiment of the present invention.
In a specific embodiment, the invention provides a method for overhauling power plant equipment, which comprises the following steps:
s11: acquiring maintenance record data of the power plant equipment in real time according to a preset period, wherein the maintenance record data comprises equipment, components, fault modes, maintenance duration and fault time;
the method comprises the steps that interface programs such as enterprise asset management can be collected according to different periods, the data comprise equipment maintenance record data, equipment defect data, work ticket data, equipment power generation plan data, equipment maintenance plan data and the like, all the maintenance record data are judged reasonably according to equipment, components, fault modes, maintenance duration and fault time, and repeated and invalid data are removed;
and for the data meeting the conditions, structured maintenance record data is formed according to equipment, components, fault modes, maintenance duration and fault time, a historical equipment fault sample database is formed, and the historical fault sample database of the equipment of the same type can be referred to for the newly-built unit.
S12: processing the data of the maintenance record to obtain a distribution function F of the failure moden(X), wherein n is the number of failure modes, and X is a vector consisting of equipment, components, a maintenance mode, maintenance time and failure time; for example, the distribution function of the failure mode can be obtained through a machine learning algorithm, such as a neuron algorithm or a random forest algorithm.
S13: processing the data of the maintenance record to obtain a distribution function f of reliabilityn(t), where n is the number of failure modes and t is time;
the overhaul record data can be processed by a statistical analysis and data fitting method, and the reliability distribution rule of the fault mode along with time is obtained.
S14: respectively carrying out distribution function F on the fault modes according to the obtained influence coefficients of the running environment and the state of the equipment on the reliability of the fault modesn(X) and distribution function f of reliabilityn(t) correcting to obtain the distribution function F of the corrected failure modesn' (X) and distribution function f of corrected reliabilityn'(t);
Obtaining an influence coefficient of an equipment operation environment and a state on the reliability of an equipment fault mode, wherein the equipment operation environment comprises an environment where the equipment is located, such as indoor or outdoor environment, ambient temperature and the like; the state characteristics include the operating parameters of the equipment such as load, vibration and temperature.
S15 obtaining a loss function L generated after the equipment failure mode occursnWherein n is the number of failure modes; the losses include equipment losses and economic losses due to less power generation after equipment outage.
S16 according to the loss function L
nAnd distribution function F of corrected failure modes
n' (X) obtaining a first function of economic loss probability distribution
S17 according to the loss function L
nAnd distribution function f of corrected reliability
n' (t) obtaining a second function of economic loss probability distribution
S18 finding a first function L of the probability distribution of economic losses by a loop iteration methodMRelative first minimum equipment failure loss value CM;
S19: according to the input maintenance strategy data, a second function l is given to the probability distribution of economic lossRThe data processing is carried out to obtain an equipment fault loss distribution function LRAnd seeking a second minimum equipment fault loss value C by a loop iteration methodR;
S20: calculating r-C according to the same maintenance strategyR/CMAnd stopping iteration when r is 0.9-1.1, and finally, setting the maintenance strategy as the current first minimum equipment fault loss value CMAnd a second minimum equipment failure loss value CRAnd (4) any corresponding maintenance strategy is used for maintaining the power plant equipment by utilizing the final maintenance strategy.
And when the r value does not meet the condition, changing the initial input value of the maintenance strategy until the condition is met.
By applying the method and the system for overhauling the power plant equipment, provided by the invention, the data of the overhauling record data is processed to obtain the distribution function of the failure mode, and the loss function generated after the failure mode occurs is combined with the distribution function of the failure mode to form a first function of the economic loss probability distribution to obtain a first minimum equipment failure loss value; the data of the maintenance record data are processed to obtain a distribution function of reliability, and a loss function generated after the fault mode occurs is combined with the distribution function of reliability to form a second function of economic loss probability distribution to obtain a second minimum equipment fault loss value; and stopping iteration by enabling the first minimum equipment fault loss value and the second minimum equipment fault loss value to meet a certain ratio, wherein the final overhaul strategy is any one detection strategy corresponding to the current first minimum equipment fault loss value and the current second minimum equipment fault loss value, and the final overhaul strategy is used for overhauling the power plant equipment.
The method combines the maintenance strategy, the fault and the maintenance task, combines the reliability distribution function of the equipment fault mode with the economic loss when the fault mode occurs, obtains the economic loss probability distribution function of the equipment, converts the reliability distribution of the equipment into the economic loss distribution of the equipment fault, and meets the requirements of users.
Specifically, according to the input overhaul strategy data, a second function l is distributed to the economic loss probabilityRThe data processing is carried out to obtain an equipment fault loss distribution function LRAnd seeking a minimum equipment fault loss value C by a loop iteration methodRPreviously, the method further comprises:
obtaining the future of the unitPlanned power generation
Wherein p (t) is a function of the power and time of the unit, t1For optimum start time of overhaul, t2For the maintenance end time, the maintenance time length T is T ═ T2-t1;
Best start time t for overhaul
1Planning the generation of electricity for the future of a crash
The time corresponding to the minimum;
the overhaul strategy data comprises equipment, components, overhaul modes and overhaul duration T.
The future planned generating capacity of the unit is an integral value of a function p (t) of the unit power and time in a specified time range, and the optimal start time t of maintenance
1Should satisfy
Further, a second function l for the economic loss probability distributionRThe data processing is carried out to obtain an equipment fault loss distribution function LRThe method specifically comprises the following steps:
for the economic loss probability distributionRObtaining an equipment fault loss distribution function L based on overhaul policy data by a Monte Carlo simulation methodR。
For the economic loss probability distributionRAcquiring an equipment fault loss distribution function L obtained based on equipment, components, maintenance modes and maintenance duration by adopting a Monte Carlo simulation methodRSeeking a second minimum equipment fault loss value C from a plurality of maintenance strategies through a loop iteration methodR。
Specifically, the data processing is carried out on the overhaul record data to obtain a distribution function F of the fault moden(X), specifically including:
classifying the historical samples according to six attributes of equipment, components, fault modes, maintenance duration and fault time; for example: the equipment attributes may include an induced draft fan, a feed pump, a steam turbine, a condensate pump and the like; the parts may include bearings, bushings, mechanical seals, oil coolers, etc.; failure modes may include bearing vibration anomalies, bearing temperature anomalies, mechanical seal leakage, oil cooler fouling, etc.; the mode of maintenance may include inspection, maintenance, component replacement, etc.; the overhaul duration is the time taken to eliminate the determined failure mode and can be divided by hours; the failure time is the calendar time when the failure mode of the device occurred.
Taking the distribution rule of the fault mode as a problem to be solved, and respectively taking five attributes of equipment, components, a maintenance mode, maintenance duration and fault time as root nodes to construct a decision tree of the distribution rule of the fault mode;
respectively calculating information gains of five attributes of equipment, components, maintenance mode, maintenance duration and fault time to obtain a distribution function F of a fault moden(X)。
Or in one embodiment, the data processing is carried out on the maintenance record data to obtain a distribution function F of the failure moden(X), specifically including:
classifying the historical samples according to six attributes of equipment, components, fault modes, maintenance duration and fault time;
respectively calculating information gains of five attributes of equipment, components, a maintenance mode, maintenance duration and fault time, and taking a decision tree constructed by the attribute with the maximum information gain as a decision tree of a fault mode distribution rule;
acquiring a distribution function F of a failure mode by adopting an integrated learning algorithm on the constructed decision tree with five attributesn(X)。
Here and above decision tree algorithm may adopt decision tree algorithms such as D3, C4.5, etc., and selection of the decision tree algorithm may be performed according to actual needs, and an integrated learning algorithm such as an averaging method, a voting method, a learning method, etc. is integrated. The number of n depends on the number of failure modes, and X is a vector consisting of 5 attributes of equipment, components, a maintenance mode, maintenance duration and failure time.
Further, data processing is carried out on the overhaul record data to obtain a distribution function f of reliabilityn(t), specifically including:
carrying out Weibull distribution/normal distribution/Gaussian distribution on the maintenance record data for data fitting to obtain a distribution function f of reliabilityn(t)。
Based on the above method embodiment, the present invention further provides a maintenance system of a power plant device, which is compared with the above method embodiment, and the system includes:
the maintenance record data acquisition module is used for acquiring maintenance record data of the power plant equipment in real time according to a preset period, wherein the maintenance record data comprises equipment, components, a fault mode, a maintenance mode, maintenance time and fault time;
a failure mode distribution function processing module for processing the data of the maintenance record to obtain a failure mode distribution function Fn(X), wherein n is the number of failure modes, and X is a vector consisting of equipment, components, a maintenance mode, maintenance time and failure time;
a reliability distribution function processing module for processing the data of the maintenance record to obtain a reliability distribution function fn(t); wherein n is the number of failure modes, and t is time;
an influence coefficient correction module for respectively performing distribution function F on the fault mode according to the obtained influence coefficients of the equipment operating environment and the state on the reliability of the fault moden(X) and distribution function f of reliabilityn(t) correcting to obtain the distribution function F of the corrected failure modesn' (X) and distribution function f of corrected reliabilityn'(t);
A loss function obtaining module for obtaining a loss function L generated after the device failure mode occursnWherein n is the number of failure modes;
an economic loss probability distribution first function processing module for processing the economic loss probability distribution according to the loss function L
nAnd distribution function F of corrected failure modes
n' (X) obtaining a first function of economic loss probability distribution
A second function processing module for economic loss probability distribution according to the loss function L
nAnd distribution function f of corrected reliability
n' (t) obtaining a second function of economic loss probability distribution
A first minimum equipment failure loss value calculation module for searching a first function L related to the economic loss probability distribution by a loop iteration methodMRelative first minimum equipment failure loss value CM;
A second minimum equipment fault loss value calculation module for calculating a second function l of the economic loss probability distribution according to the input maintenance strategy dataRThe data processing is carried out to obtain an equipment fault loss distribution function LRAnd seeking a second minimum equipment fault loss value C by a loop iteration methodR;
A ratio calculation module for calculating r ═ C according to the same maintenance strategyR/CMAnd stopping iteration when r is 0.9-1.1, and finally, setting the maintenance strategy as the current first minimum equipment fault loss value CMAnd a second minimum equipment failure loss value CRAnd (4) any corresponding maintenance strategy is used for maintaining the power plant equipment by utilizing the final maintenance strategy.
Specifically, the system further comprises:
the maintenance time calculation module is used for acquiring the future planned generated energy of the unit
Wherein p (t) is a function of the power and time of the unit, t
1For optimum start time of overhaul, t
2For the maintenance end time, the maintenance time length T is T ═ T
2-t
1(ii) a Best start time t for overhaul
1Planning the generation of electricity for the future of a crash
The time corresponding to the minimum; the overhaul strategy data comprises equipment, components, overhaul modes and overhaul duration T.
Further, the second minimum device fault loss value calculation module is specifically configured to:
distribution function l for equipment fault lossRObtaining an equipment fault loss distribution function L based on overhaul policy data by a Monte Carlo simulation methodR。
Further, the failure mode distribution function processing module is specifically configured to:
classifying the historical samples according to six attributes of equipment, components, fault modes, maintenance duration and fault time;
taking the distribution rule of the fault mode as a problem to be solved, and respectively taking five attributes of equipment, components, a maintenance mode, maintenance duration and fault time as root nodes to construct a decision tree of the distribution rule of the fault mode;
respectively calculating information gains of five attributes of equipment, components, maintenance mode, maintenance duration and fault time to obtain a distribution function F of a fault moden(X)。
The present invention also provides an apparatus comprising a memory and a processor, wherein:
a memory for storing a computer program;
a processor for executing a computer program for implementing a method of servicing a power plant as in any one of the embodiments described above.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.