CN112528519A

CN112528519A - Method, system, readable medium and electronic device for engine quality early warning service

Info

Publication number: CN112528519A
Application number: CN202011566228.9A
Authority: CN
Inventors: 刘尚文; 谢宽; 赵盛轩; 张力佳; 朱丹; 元志林; 杨本晓; 刘夏青
Original assignee: Beijing Saibao Industrial Technology Research Institute Co ltd
Current assignee: Beijing Saibao Industrial Technology Research Institute Co ltd
Priority date: 2020-05-14
Filing date: 2020-12-25
Publication date: 2021-03-19
Also published as: CN111597708A

Abstract

The invention relates to a method, a system, a readable medium and an electronic device for engine quality early warning service, wherein the method comprises the following steps: constructing a correlation analysis data table according to the detection data of the sensor acquired from the historical fault data and the corresponding fault type; modeling an association rule of the engine fault to obtain a frequent item set on an association analysis data table; constructing an FP tree according to the correlation analysis data table and the minimum support degree of the engine fault; excavating a frequent item set for the FP tree by utilizing an FP-Growth algorithm; and sequencing the frequent item sets obtained by mining in the sets according to the indexes to obtain a potential fault set. The invention can find potential faults which are not found but possibly exist temporarily on the basis of the found faults of the engine, can mine the potential faults of the engine according to the detection data of the sensor in real time, and can actually prolong the service life of the engine by analyzing the potential faults and repairing the potential faults in time, thereby improving the quality of after-sale service.

Description

Method, system, readable medium and electronic device for engine quality early warning service

Cross-referencing

The invention requires the application numbers submitted at 14/05/2020 as: 202010409035.6, the priority of the invention to the chinese patent application entitled "method, system, readable medium, and electronic device for compressor quality alert service", the disclosure of which is hereby incorporated by reference into this application.

Technical Field

The invention relates to the technical field of engine quality monitoring, in particular to a method, a system, a readable medium and electronic equipment for engine quality early warning service.

Background

Engines, an industrial common electromechanical device, are often subject to various types of failures, common failure types include but are not limited to piston ring leakage, valve spring looseness or damage, insufficient cooling water supply, loose or broken coupling, air intake leakage, and the like.

The conventional fault detection method is usually established on the basis of observing and analyzing various signal data of an engine monitored in real time, and according to engineering practice experiences, signals indicating that the engine possibly fails comprise oxygen sensor faults (detecting oxygen sensor voltage, engine rotating speed, air flow value and the like), throttle valve controller signal faults (detecting air flow, throttle valve opening, oil injection compensation amount and the like), ignition ring faults (detecting ignition advance angle and the like), knocking or knocking (detecting engine rotating speed, oil injection pulse, oil pump switch and the like) and the like. Through the precise arrangement of sensors, the monitoring signals can be embodied on parts of the engine to a certain extent, and the available parameters comprise coolant temperature, engine load, engine speed, throttle angle, battery voltage, idle air flow value and the like, but the existing fault detection methods determine whether a fault exists according to the results shown by the fault and cannot effectively warn about the potential fault.

Based on the above, there is a problem in the prior art that a potential fault of the engine cannot be excavated based on the detection data.

The above drawbacks are expected to be overcome by those skilled in the art.

Disclosure of Invention

Technical problem to be solved

In order to solve the above problems in the prior art, the present invention provides a method, a system, a readable medium and an electronic device for providing an engine quality early warning service, so as to overcome, at least to some extent, the problem in the prior art that a potential fault of an engine cannot be detected.

(II) technical scheme

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

according to a first aspect of the present invention, there is provided a method of an engine quality warning service, comprising:

constructing a correlation analysis data table according to the detection data of the sensor acquired from the historical fault data and the corresponding fault type;

modeling an association rule of the engine fault to obtain a frequent item set on an association analysis data table;

constructing an FP tree according to the correlation analysis data table and the minimum support degree of the engine fault;

excavating a frequent item set for the FP tree by utilizing an FP-Growth algorithm;

and sequencing the frequent item sets obtained by mining in the sets according to the indexes to obtain a potential fault set.

In an embodiment of the present invention, constructing the association analysis data table according to the detection data of the sensor obtained from the historical fault data and the corresponding fault type includes:

constructing a detection data set based on detection data of the historical fault data acquisition sensor, wherein x_i＝[x_i,1,x_i,2,…,x_i,n]Representing the detection data of a sensor of an ith engine under n different characteristics, wherein n represents the number of characteristics, i belongs to {1, …, m }, and the characteristics comprise at least two of coolant temperature, engine load, engine speed, throttle angle, storage battery voltage and idle air flow value;

obtaining fault type y of ith engine based on historical fault data_i，y_iDifferent fault types are indicated by different values;

determining a detection sample according to the detection data set and the fault type;

processing the detection sample to obtain a processed detection sample;

generating a transaction for correlation analysis based on the processed detection sample in combination with a preset threshold;

constructing a correlation analysis data table by a plurality of transactions for correlation analysis, and marking as

In the association analysis data table

T_i＝{y_i}。

In an embodiment of the present invention, the generating a transaction for association analysis based on the processed detection samples and the preset threshold includes:

initializing m-1 sets, each set containing the fault type of the corresponding sample,

T_i＝{y_i}；

determining an upper error bound as ∈ for each feature_jWhere j ∈ {1, …, n };

randomly selecting a first sample and a second sample from all detection samples in the detection data set, and comparing the first sample and the second sample in pairs to calculate a confidence ratio;

judging two detection samples with characteristics meeting the upper error limit and the confidence ratio being greater than a preset threshold value as similar samples;

adding the fault type corresponding to the second sample in the similar samples to the set T of the first sample_iIn (1), a transaction for association analysis is obtained.

In an embodiment of the present invention, modeling the association rule of the engine fault to obtain a frequent item set on the association analysis data table includes:

record the set of all items in the association analysis data table as

Determining the support degree of the engine correlation fault according to the correlation analysis data table

And is

Will "rule T_p→T_kThe support degree of (2) is denoted as support (T)_p→T_k) The calculation method is as follows:

support(T_p→T_k)＝P_D(T_p∪T_k)

wherein P is_D(x) Indicating that transaction x is issued in association analysis data table DThe frequency of the generation;

determining confidence of engine associated fault according to support degree, and converting' rule T_p→T_kThe support degree of (c) is denoted as confidence (T)_p→T_k) The calculation method is as follows:

setting the minimum values of the support degree and the confidence degree to be respectively recorded as the minimum support degree s_minAnd minimum confidence c_minIf support (T)_p→T_k)≥s_minAnd confidence (T)_p→T_k)≥c_minThen T is_p→T_kIs a strong association rule;

obtaining a frequent item set of the engine association fault based on the confidence degree and the strong association rule, and setting

If T_pThe probability of occurrence in the correlation analysis data table D is greater than or equal to s_min(m-1), then the obtained T_pAnalyzing a set of frequent items, T, in a data sheet for associations_p＝{y_i,1,y_i,2,…,y_i,q}。

In an embodiment of the invention, the building the FP-tree according to the correlation analysis data table and the minimum support degree of the engine fault comprises:

comparing the frequent item sets in the association analysis data table one by using the minimum support degree, and deleting the items with the support degree smaller than the minimum support degree in the association analysis data table to obtain a set L containing all frequent item sets, wherein the frequent item set is a set with only one frequent item;

sorting all frequent item sets in the set L from large to small according to the support degree;

creating a root node of the FP tree, and ordering the transaction T 'at the front'_iLinking to the root nodes in sequence, merging the parts sharing the same root node, and adding 1 to the numerical value corresponding to the shared item;

and if the new node exists after the insertion, linking the node corresponding to the item head table to the new node through the node link table to obtain the FP tree.

In an embodiment of the present invention, mining a frequent item set for a FP-tree by using a FP-Growth algorithm includes:

establishing an item header, wherein the last node of each row stores a pointer pointing to a corresponding item in the FP tree, and each row is a linked list, wherein the item header is a set of the linked lists;

finding a prefix path for each frequent item set to form a conditional mode base;

sequentially finding a conditional mode base corresponding to the item header from the bottom item of the item header upwards, and establishing a conditional FP tree;

recursively mining a frequent item set of an item head table from a conditional mode base for the conditional FP tree;

judging whether the number of the items of the frequent item set is limited, and if not, directly returning to the frequent item set obtained by mining; if limited, only the frequent set of items that meet the number of items requirement is returned.

In an embodiment of the present invention, the sorting the frequent item sets obtained by mining according to the indexes in the sets, and obtaining the potential fault set includes:

sequencing each frequent item set obtained by mining in a set according to indexes, and respectively recording the obtained frequent item sets as F₁，...，F_k；

Initializing length vector c_iIs a zero vector, c_iIs a vector of length | l | -1;

defining a strong rule threshold value in the association rule as s;

from F₁，...，F_kIn turn select an item F_jJudging the type of the fault y_iWhether or not it belongs to F_jIf y is_iBelong to F_jThen for F_jMiddle division of y_iAny other element y_p

Length vector c

_i1 is added to the element at the p-th position of (1);

if the length vector c_iGreater than or equal to the strong rule threshold s, then the frequent item set is merged intoAnd obtaining a potential fault set corresponding to each obvious fault in the frequent item set in the potential fault set.

According to a second aspect of the present invention, there is also provided a system for an engine quality warning service, comprising:

the data table module is used for constructing a correlation analysis data table according to the detection data of the sensor acquired from the historical fault data and the corresponding fault type;

the rule modeling module is used for modeling the association rule of the engine fault to obtain a frequent item set on the association analysis data table;

the FP tree building module is used for building the FP tree according to the correlation analysis data table and the minimum support degree of the engine fault;

the data mining module is used for mining a frequent item set for the FP tree by utilizing an FP-Growth algorithm;

and the potential fault module is used for sequencing the frequent item sets obtained by mining in the sets according to the indexes to obtain potential fault sets.

According to a third aspect of the present invention, there is also provided an electronic apparatus comprising:

a processor;

a memory storing instructions for the processor to control the method steps described above.

According to a fourth aspect of the present invention, there is also provided a computer-readable medium having stored thereon computer-executable instructions which, when executed by a processor, implement the method steps described above.

(III) advantageous effects

The invention has the beneficial effects that: according to the method, the system, the readable medium and the electronic device for the engine quality early warning service, provided by the embodiment of the invention, after the FP tree is constructed according to the correlation analysis data sheet and the minimum support degree of the engine faults, the frequent item set is mined for the FP tree through the application of the FP-Growth algorithm, potential faults which are not found temporarily but possibly exist can be found on the basis of the found engine faults, the potential faults of the engine can be obtained through mining according to the detection data of the sensor in real time, and the potential faults are analyzed and repaired in time, so that the service life of the engine is practically improved, and the quality of after-sale services is improved.

Drawings

FIG. 1 is a flow chart of a method of providing engine quality warning service according to one embodiment of the present invention;

FIG. 2 is a flowchart illustrating step S110 in FIG. 1 according to an embodiment of the present invention;

FIG. 3 shows T 'in the construction of FP tree in an embodiment of the present invention'_iSchematic diagram inserted into FP tree;

FIG. 4 shows T 'in the construction of FP tree in an embodiment of the present invention'₂Schematic diagram inserted into FP tree;

FIG. 5 shows merged T 'in FP tree construction in an embodiment of the present invention'₁And T'₂A schematic diagram of the shared prefix of (a);

FIG. 6 is a diagram illustrating a final FP tree in accordance with an embodiment of the present invention;

FIG. 7 is a diagram illustrating the establishment of an entry header table based on the FP tree shown in FIG. 6 according to an embodiment of the present invention;

FIG. 8 shows { y } in an embodiment of the present invention₄Schematic of conditional FP-tree of h;

FIG. 9 is a block diagram of an embodiment of the present invention for updating y₄Schematic of conditional FP-tree of h;

FIG. 10 shows { y } in an embodiment of the present invention₅-a schematic representation of the final conditional FP-tree;

FIG. 11 is a schematic diagram of a system for providing engine quality warning services in accordance with another embodiment of the present invention;

fig. 12 is a schematic structural diagram illustrating a computer system of an electronic device according to an embodiment of the present invention.

Detailed Description

For the purpose of better explaining the present invention and to facilitate understanding, the present invention will be described in detail by way of specific embodiments with reference to the accompanying drawings.

All technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

The existing fault detection method generally takes each monitoring data as the input of a classification model, takes the fault type as the output of the classification model, and trains model parameters based on the existing database samples, so that a certain fault corresponding to the state can be given to any given monitoring data input. The fault detection methods belong to supervised learning, and the adopted models comprise a support vector machine, an artificial neural network and the like, but the methods cannot find other potential faults of the engine. For example, when it is found that the engine speed is too high, the fault detection method may possibly draw a conclusion that the engine throttle control signal is faulty with the injector (because of vacuum leakage or resulting in increased fuel injection), but in reality the fault that the engine speed is too high may be associated with the oxygen sensor fault, and the "potential" associated fault is not recognized by the conventional machine learning fault detection method.

Based on the above, the invention provides a method, a system, a readable medium and electronic equipment for engine quality early warning service based on FP-Growth algorithm, so as to solve the problems of quality monitoring and fault early warning of motorcycle engines.

Fig. 1 is a flowchart of a method for providing an engine quality early warning service according to an embodiment of the present invention, as shown in fig. 1, specifically including the following steps:

as shown in fig. 1, in step S110, an association analysis data table is constructed according to the detection data of the sensor obtained from the historical fault data and the corresponding fault type;

as shown in fig. 1, in step S120, modeling is performed on the association rule of the engine fault, so as to obtain a frequent item set on the association analysis data table;

as shown in fig. 1, in step S130, a FP-tree is constructed according to the correlation analysis data table and the minimum support degree of the engine fault;

as shown in fig. 1, in step S140, a FP-Growth algorithm is used to mine a frequent item set for the FP-tree;

as shown in fig. 1, in step S150, the frequent item sets obtained by mining are sorted by indexes within the set to obtain a latent fault set.

According to the method, in order to find out potential engine faults, a calculation method of the incidence relation among various faults of the engine is provided based on the FP-Growth algorithm, and therefore early warning of the potential faults is provided according to the incidence rule on the basis of a supervised learning algorithm. Therefore, the device management personnel can timely find and repair potential faults which are not found by the supervised learning model, and therefore the efficiency of the engine quality early warning comprehensive service system and the quality of after-sales service are improved.

The steps of the method shown in fig. 1 are described in detail below:

in step S110, an association analysis data table is constructed from the detection data of the sensors acquired from the historical failure data and the corresponding failure types.

In an embodiment of the present invention, in this step, the detection sample is mainly determined according to the existing fault detection record, and is processed into data in the form of an association analysis data table.

Fig. 2 is a flowchart of step S110 in fig. 1 according to an embodiment of the present invention, as shown in fig. 2, which specifically includes the following steps:

step S201, a detection data set is constructed based on the detection data of the historical fault data acquisition sensor.

Assuming that a factory has deployed temperature, air pressure and other sensor devices on m engines,

wherein x in the historical fault data_i＝[x_i,1,x_i,2,…,x_i,n]Represents the detected data of the sensor of the ith engine under n different characteristics, wherein n represents the number of characteristics, i is equal to {1, …, m }, wherein the characteristics comprise the temperature of the cooling liquid, the load of the engine, the rotating speed of the engine, the angle of a throttle valve, the voltage of a storage battery and the idle airflowAt least two of the magnitudes.

The historical fault data comprises that the voltage of an oxygen sensor is too high or too low, the rotating speed of an engine is too high or too low, the idle air flow value is too high or too low, the opening of a throttle valve is large or small, an oil pump is abnormal in opening and closing, an oil injection pulse is abnormal, an oil injection compensation quantity is abnormal, the angle of the throttle valve is too high, an ignition advance angle is abnormal and the like, fault early warning is carried out through big data mining and correlation analysis, and the types of the early warning include but are not limited: the method comprises the following steps of oxygen sensor fault early warning, throttle controller early warning, ignition ring fault early warning, engine overload early warning, oil injector fault early warning, cylinder knocking or explosion early warning or other early warnings and the like.

Step S202, obtaining the fault type y of the ith engine based on historical fault data_i，y_iDifferent fault types are indicated by different values. y is_iIndicating the type of fault found by the i-th engine through manual inspection or a deployed supervised learning model, e.g. y _i1 represents oxygen sensor failure, y _i2 represents a fuel injector failure fault, etc.

Step S203, determining the detection sample according to the detection data set and the fault type, for example, { x ] may be used in this embodiment_i，y_iDenotes a test sample.

And step S204, processing the detection sample to obtain a processed detection sample.

After the detection sample is obtained, normalization processing is firstly carried out on detection data in the detection sample.

Calculating a sample mean for each column of features

x_i，jHistorical sensed data representing a jth characteristic of an ith engine;

then, the sample standard deviation for each column of observed features is calculated:

the normalized detection sample is obtained by calculation according to the following formula:

step S205, generating a transaction for correlation analysis based on the processed detection sample and a preset threshold.

In this step

Setting the upper limit of the error degree of the jth characteristic as ∈_jThe initial value of the count is 0, and whether the two samples can be classified as 'similar samples' is judged according to a 70% confidence ratio, and the specific operation is carried out according to the following steps:

based on 1) -9) above, initializing m-1 sets, each set containing the fault type of the corresponding sample,

T_i＝{y_i}; determining an upper error bound as ∈ for each feature_jWhere j ∈ {1,..., n }; randomly selecting a first sample and a second sample from all detection samples in the detection data set, and comparing the first sample and the second sample in pairs to calculate a confidence ratio; judging two detection samples with characteristics meeting the upper error limit and the confidence ratio being greater than a preset threshold value as similar samples; adding the fault type corresponding to the second sample in the similar samples to the set T of the first sample_iIn (1), get used for correlationThe transaction is analyzed. And judging whether the whole observation samples are classified into 'similar samples' pairwise according to whether the two samples have enough characteristics within a given error upper limit. If count is greater than or equal to 70% · (m-1), then

Indicating that if the latter and former are "similar", the fault type corresponding to the latter is added (apend) to the set T of the former.

In step S206, an association analysis data table is constructed by a plurality of transactions for association analysis.

At this point, a data table of user correlation analysis is constructed for the fault finding problem of the engine and is recorded as

In the association analysis data table

T_i＝{y_i}. Each behavior T of the data table_iIt is considered as one containing | T_iA transaction of | items. Each entry is a type of engine fault, and there are m-1 transactions in the correlation analysis data table.

In step S120, a modeling is performed on the association rule of the engine fault to obtain a frequent item set on the association analysis data table.

The data table obtained in step S110 is denoted as

Record the set of all items in the association analysis data table as

Definition 1: determining the support degree of the engine correlation fault according to the correlation analysis data table

And is

support(T_p→T_k)＝P_D(T_p∪T_k) Formula (4)

Wherein P is_D(x) Represents the frequency with which transaction x occurs in the correlation analysis data table D;

definition 2: determining confidence coefficient of engine associated fault according to support degree, and setting

And is

Will "rule T_p→T_kThe support degree of (c) is denoted as confidence (T)_p→T_k) The calculation method is as follows:

definition 3: strong association rule, let

And is

definition 4: obtaining a frequent item set (frequency item sets) of the engine association fault based on the confidence degree and the strong association rule, and setting

If T_pThe probability of occurrence in the correlation analysis data table D is greater than or equal to s_min(m-1), then the obtained T_pAnalyzing a set of frequent items, i.e. T, in a data sheet for associations_p＝{y_i,1,y_i,2,…,y_i,q}。

Based on step S120, those meaningless rules are removed by defining a support degree, and defining a confidence degree makes the inference reliable. As long as all the frequent item sets are found, potential engine faults that have not been directly detected can be mined. E.g. T_p＝{y_i,1,y_i,2,…,y_i,qThe method is a frequent item set, and for the ith engine, only the fault type y can be found according to manual detection or a deployed supervised learning model_i,1. But with a frequent itemset T_pThen, other faults y which may exist on the ith engine can be found_i,2,…,y_i,qFactory-based Hadoop big data platform will be simultaneously aligned to y_i,1And y_i,2,…,y_i,qThis type of fault alerts and thus provides a comprehensive troubleshooting of the potential fault. Therefore, the next step mainly utilizes FP-Growth algorithm to mine all the frequent item sets on the engine correlation analysis data table D.

In step S130, the FP-tree is constructed according to the correlation analysis data table and the minimum support degree of the engine failure.

In the step, the data table D and the minimum support degree s are mainly analyzed according to the input association_minConstructing an FP tree and mining a frequent item set, and specifically comprising the following steps:

firstly, performing time-to-time comparison on frequent item sets in the association analysis data table by using the minimum support degree, and deleting the items with the support degree smaller than the minimum support degree in the association analysis data table to obtain a set L containing all frequent item sets, wherein the frequent item set is a set with only one frequent item, and the frequent n item set is a set containing n frequent items. In the step, the read original data is removed from the non-frequent item set by taking the minimum support degree as a condition after the data table is scanned, and the obtained frequent item set meets the requirements of only one item and the minimum support degree.

Secondly, all the frequent item sets in the set L are sorted from large to small according to the support degree.

Then, a root node of the FP tree is created, and the transactions T 'which are sequenced at the front are sequenced'_iAnd linking to the root nodes in sequence, merging the parts sharing the same root node, and adding 1 to the numerical value corresponding to the shared item. In this step, the sorted data set is read in and inserted into the FP-tree. And when inserting, inserting into the FP tree according to the sequence after sequencing, wherein the node in the front of the sequence is an ancestor node, and the node in the back of the sequence is a descendant node. If there is a common ancestor, the corresponding common ancestor node count is incremented by 1.

And finally, if a new node exists after the insertion, linking the node corresponding to the item head table to the new node through the node link table to obtain the FP tree. After inserting the FP tree in this step, if a new node appears, the node corresponding to the entry head table may be linked to the new node through the node linked list, and the establishment of the FP tree is completed until all data are inserted into the FP tree.

In step S140, a FP-Growth algorithm is used to mine a frequent item set for the FP tree.

The FP tree is used for quickly generating frequent item sets, records in the database are inserted into one tree according to a certain sequence, and the FP tree can compress a plurality of records in one path, so that the search time can be shortened, and the search efficiency can be improved. In the step, the FP tree is further mined by adopting an algorithm, and the specific steps are as follows:

firstly, establishing an item header, storing a pointer pointing to a corresponding item in the FP tree by the last node of each row, and storing a linked list by each row, wherein the item header is a set of the linked lists; secondly, searching a prefix path for each frequent item set to form a conditional mode base; then, sequentially finding a conditional mode base corresponding to the item header from the bottom item of the item header upwards, and establishing a conditional FP tree; then, recursively mining a frequent item set of an item head table from a conditional mode base for the conditional FP tree; finally, judging whether the number of terms of the frequent item set is limited or not, and if the number of terms of the frequent item set is not limited, directly returning to the frequent item set obtained by mining; and if the number of items of the frequent item set is limited, only returning the frequent item set meeting the requirement of the number of items.

In step S150, the frequent item sets obtained by mining are sorted according to the index in the set, so as to obtain a potential failure set.

In this step, by merging frequent item sets, a potential failure set is found, which specifically includes:

firstly, sequencing each frequent item set obtained by mining in a set according to indexes, and respectively recording the obtained frequent item sets as F₁，...，F_k；

Next, the length vector c is initialized_iIs a zero vector, c_iIs a vector of length | l | -1;

then, defining a strong rule threshold value in the association rule as s;

then, from F₁，...，F_kIn turn select an item F_jJudging the type of the fault y_iWhether or not it belongs to F_jIf y is_iBelong to F_jThen for F_jMiddle division of y_iAny other element y_pLength vector c_iPlus 1 for the element in the p-th position. Due to y_pIs represented by F_jMiddle y_iAny other fault type, y_pP in (1) represents the number of this element, according to which p can be in the length vector c_iThe correct position is indexed internally, and in the step, all the attributes F are subjected to traversal operation_j-{y_iY of_pCorresponding counter c_i(p) are all self-increasing by 1;

finally, for the length vector c_iIf the current element value is greater than or equal to the strong rule threshold s, identifying the frequent item corresponding to the element value as y_iIs detected. Thus, y can be obtained_iIs detected.

The following specifically describes steps S130 to S150 with reference to specific examples:

(1) set I of frequent item sets as { y₁，y₂，y₃，y₄，y₅In which y is₁To y₅Are respectively provided withIndicating a type of excavator malfunction such as valve spring damage, insufficient cooling water, broken coupling, overload operation, piston seizure, etc. Let m be 9, and the correlation analysis data table D constructed according to step S110 is shown in table 1 below:

TABLE 1

(2) Global scanning one pass through the associative analysis data table D, finding a "frequent item set" (where an item set is a set of only one item and a frequent item set is a set of items satisfying the minimum support) according to the minimum support. According to the scanning result, all the item sets are ordered according to the corresponding support degree to obtain { { y₂：7}，{y₁：6}，{y₃：6}，{y₄：2}，{y₅: 2}}. With a minimum support of 2, i.e. s_minAll sets of items are one frequent item set, and the set of all frequent item sets is denoted as L.

(3) And creating a root node of the FP tree, and marking as null. And sequencing each transaction data in the database according to the sequence of the frequent item set, storing the transaction data in the FP pattern tree, and establishing an item head table. Specifically, the method comprises the following steps: each transaction in the association analysis data table D is inserted into the FP-tree. If there is a common ancestor node, then the corresponding common ancestor node count is incremented by 1. After insertion, if a new node appears, the node corresponding to the entry head table is linked with the new node through the node linked list. And completing the building of the FP tree until all data are inserted into the FP tree. The count on the node is used for subsequent judgment corresponding to the support degree of the current item set, and the method specifically comprises the following steps:

(3.1) for each transaction T according to the binding L_iAnd (6) sorting. By T₁To give example, T'_i＝{{y₂：1}，{y₁：1}，{y₅：1}}。T₂To T₉And so on.

(3.2) ordering each of the ordered transactions T'_iWhich in turn links to the FP-tree. FIG. 3 shows T 'in the construction of FP tree in an embodiment of the present invention'_iSchematic representation of insertion into FP Tree, as shown in FIG. 3, for T'₁Will y is₂、y₁、y₅In turn linked to the root node. FIG. 4 shows T 'in the construction of FP tree in an embodiment of the present invention'₂Schematic diagram of inserting into FP tree, as shown in FIG. 4. T 'in figure 4'₂And T'₁Shared prefix path null → y₂Therefore, the shared parts are merged, and the value corresponding to the shared item is added by 1, fig. 5 is merged T 'in FP-tree construction according to an embodiment of the present invention'₁And T'₂As shown in fig. 5, will share the prefix y₂After merging, the corresponding value is added with 1 to be 2, namely y₂: 2. likewise, for subsequent T'₃To T'₉If the prefix path is the same as the prefix path of the already-linked transaction on the current FP-tree, merging, and adding 1 to the corresponding node count, where fig. 6 is a schematic diagram of the FP-tree finally obtained in an embodiment of the present invention.

And (3.3) establishing an item head table, wherein the last node of each row stores a pointer pointing to a corresponding item in the FP tree. Fig. 7 is a schematic diagram of creating an entry header table based on the FP-tree shown in fig. 6 in an embodiment of the present invention, and as shown in fig. 7, each row has a linked list, and the position of the corresponding entry in the FP-tree can be sequentially indexed.

(4) Finding a prefix path for each frequent item set, forming a conditional mode base, and establishing a 'conditional FP tree'. The entry header table is a collection of linked lists. Frequently, a set of items contains { y }₂}，{y₁}，{y₃}，{y₄And { y }₅}。

(4.1) with { y₄For example, the prefix path in the FP tree is { { y₂: 1} and { { y { { C₂：1}，{y₁: 1}}. These two prefix paths constitute y₄Is based on a conditional mode, thereby constructing { y }₄FIG. 8 shows a conditional FP tree of { y } in an embodiment of the present invention₄Schematic of conditional FP-tree of. As shown in fig. 8In this tree, because y₁: 1 is less than the minimum support, so sum { y }₄The items that constitute the frequent item set contain only y₂: 2}. Further updating the conditional FP tree, FIG. 9 is an embodiment of the present invention in which the update { y₄FIG. 9 shows a conditional FP tree containing { y }₂：2}。

(4.2) and { y }₅Taking the example, the final conditional FP-tree can be obtained according to the same procedure as (4.1). FIG. 10 shows { y } in an embodiment of the present invention₅Schematic representation of the final conditional FP-tree. The remaining frequent item set follows the same steps and is not expanded here.

(5) And recursively mining each conditional FP tree by combining the item head table to obtain a frequent item set. With { y₄For example, a conditional FP tree of { Y } can be mined to extract a frequent item set₂，y₄}; with { y₅For example, a conditional FP tree of { y } can be mined₂，y₅}、{y₁，y₅And { y }₂，y₁，y₅}. The conditional FP-tree follows the same steps and is not expanded here. Thus, a mined FP tree is obtained.

(6) Sequencing each frequent item set obtained in the steps in a set according to an index sequence, and respectively marking the obtained frequent item sets as F₁，...，F_kDefining a strong association threshold s, and then screening out a potential fault set corresponding to each item (obvious fault) according to the following steps: s₁，...，S_|I|。

To this end, for each engine i e {1, …, m }, there may be other faults y on the ith engine_i,1,y_i,2,…,y_i,q∈S_iHadoop big data of factoryThe platform will be on all latent faults y_i,1,y_i,2,…,y_i,qIn all i_qThe fault type gives an alarm to prompt the staff of the factory to carry out comprehensive troubleshooting on the potential fault.

In summary, according to the technical scheme provided by the embodiment of the invention, the data mining technology is mainly utilized, invisible potential faults are found out of the existing monitoring data, and the FP-Growth algorithm is utilized to mine the frequent item set for qualitative fault analysis. For faults in the same frequent item set, the rest faults are all qualified as potential faults except for obvious faults discovered by supervised learning. The big data analysis platform of the factory can send the obvious faults and all the potential faults to the manager at the same time, so that maintenance personnel can be dispatched quickly to carry out comprehensive investigation on the indexes of the excavator. Similar to a deep learning algorithm, the process of discovering the potential fault is based on a prediction result given after data mining, namely the fault prediction is made under the driving of the existing data under the condition of not actually contacting with the excavator equipment, and the defect that the early warning of the equipment fault is delayed by only monitoring the existing fault is overcome.

Corresponding to the above method, fig. 11 is a schematic diagram of a system for providing an engine quality warning service according to another embodiment of the present invention, and as shown in fig. 11, the system 1100 includes: a data table module 1110, a rule modeling module 1120, a FP-tree construction module 1130, a data mining module 1140, and a latent fault module 1150.

The data table module 1110 is configured to construct an association analysis data table according to the detection data of the sensor obtained from the historical fault data and the corresponding fault type; the rule modeling module 1120 is used for modeling the association rule of the engine fault to obtain a frequent item set on the association analysis data table; the FP tree construction module 1130 is used for constructing a FP tree according to the correlation analysis data table and the minimum support degree of the engine fault; the data mining module 1140 is used for mining a frequent item set for the FP-tree by using the FP-Growth algorithm; the latent fault module 1150 is configured to sort the mined frequent item sets by indexes within the set, so as to obtain a latent fault set.

The technical effects of the system adopting the engine quality early warning service provided by the embodiment of the invention are referred to the technical effects of the method, and are not repeated herein.

Referring now to FIG. 12, shown is a block diagram of a computer system suitable for use with the electronic device implementing an embodiment of the present invention. The computer system 1200 of the electronic device shown in fig. 12 is only an example, and should not bring any limitation to the function and the scope of use of the embodiments of the present invention.

As shown in fig. 12, the computer system 1200 includes a Central Processing Unit (CPU)1201, which can perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)1202 or a program loaded from a storage section 1208 into a Random Access Memory (RAM) 1203. In the RAM 1203, various programs and data necessary for system operation are also stored. The CPU1201, ROM 1202, and RAM 1203 are connected to each other by a bus 1204. An input/output (I/O) interface 1205 is also connected to bus 1204.

The following components are connected to the I/O interface 1205: an input section 1206 including a keyboard, a mouse, and the like; an output portion 1207 including a display device such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 1208 including a hard disk and the like; and a communication section 1209 including a network interface card such as a LAN card, a modem, or the like. The communication section 1209 performs communication processing via a network such as the internet. A driver 1210 is also connected to the I/O interface 1205 as needed. A removable medium 1211, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like, is mounted on the drive 1210 as necessary, so that a computer program read out therefrom is mounted into the storage section 1208 as necessary.

In particular, according to an embodiment of the present invention, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the invention include a computer program product comprising a computer program embodied on a computer-readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 1209, and/or installed from the removable medium 1211. The computer program performs the above-described functions defined in the system of the present application when executed by the Central Processing Unit (CPU) 1201.

It should be noted that the computer readable medium shown in the present invention can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present invention, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units described in the embodiments of the present invention may be implemented by software, or may be implemented by hardware, and the described units may also be disposed in a processor. Wherein the names of the elements do not in some way constitute a limitation on the elements themselves.

As another aspect, the present application also provides a computer-readable medium, which may be contained in the electronic device described in the above embodiments; or may exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs which, when executed by an electronic device, cause the electronic device to implement the method of engine quality warning service as described in the above embodiments.

For example, the electronic device may implement the following as shown in fig. 1: step S110, constructing a correlation analysis data table according to the detection data of the sensor acquired from the historical fault data and the corresponding fault type; step S120, modeling is carried out on the association rule of the engine fault to obtain a frequent item set on an association analysis data sheet; step S130, constructing an FP tree according to the correlation analysis data sheet and the minimum support degree of the engine fault; step S140, a frequent item set is mined for the FP tree by utilizing the FP-Growth algorithm; and S150, sequencing the frequent item sets obtained by mining in the sets according to the indexes to obtain a potential fault set.

It should be noted that although in the above detailed description several modules or units of the device for action execution are mentioned, such a division is not mandatory. Indeed, the features and functionality of two or more modules or units described above may be embodied in one module or unit, according to embodiments of the invention. Conversely, the features and functions of one module or unit described above may be further divided into embodiments by a plurality of modules or units.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiment of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiment of the present invention.

Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.

It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.

Claims

1. A method of engine quality warning service, comprising:

2. The method of engine quality warning service of claim 1, wherein constructing an association analysis data table based on the detection data of the sensors obtained from the historical fault data and the corresponding fault types comprises:

constructing a detection data set based on detection data of the historical fault data acquisition sensor, wherein x_i＝[x_i,1,x_i,2,…,x_i,n]The method comprises the steps that detection data of a sensor of an ith engine under n different characteristics are represented, n represents the number of the characteristics, i belongs to {1, …, m }, wherein the characteristics comprise at least two of an oxygen sensor voltage value, an engine speed, a throttle opening, an air flow value, an ignition advance angle and an oil injection pulse value;

processing the detection sample to obtain a processed detection sample;

building a gateway from multiple transactions for association analysisAnalysis data table of

In the association analysis data table

3. The method of engine quality warning service of claim 2, wherein generating a transaction for correlation analysis based on the processed detection samples in combination with a preset threshold comprises:

4. The method of engine quality early warning service as claimed in claim 3, wherein modeling the association rules for engine faults to derive a frequent item set on the association analysis data table comprises:

record the set of all items in the association analysis data table as

Determining the support degree of the engine correlation fault according to the correlation analysis data table,is provided with

And is

support(T_p→T_k)＝P_D(T_p∪T_k)

5. The method of engine quality warning service of claim 1, wherein constructing the FP-tree based on the correlation analysis data table and the minimum support for engine failure comprises:

6. The method of engine quality warning service of claim 1, wherein mining a frequent item set for a FP-tree using a FP-Growth algorithm comprises:

7. The method for engine quality warning service according to any one of claims 1 to 6, wherein the step of sorting the mined frequent item sets by index within a set to obtain a set of potential faults comprises:

sequencing each frequent item set obtained by mining in a set according to indexes, and respectively recording the obtained frequent item sets as F₁,…,F_k；

defining a strong rule threshold value in the association rule as s;

from F₁,…,F_kIn turn select an item F_jJudging the type of the fault y_iWhether or not it belongs to F_jIf y is_iBelong to F_jThen for F_jMiddle division of y_iAny other element y_pLength vector c_i1 is added to the element at the p-th position of (1);

if the length vector c_iAnd if the number of the faults is larger than or equal to the strong rule threshold s, merging the frequent item set into the potential fault set to obtain a potential fault set corresponding to each obvious fault in the frequent item set.

8. A system for engine quality warning service, comprising:

9. An electronic device, comprising:

a processor;

memory storing instructions for the processor to control the method steps of any of claims 1-7.

10. A computer-readable medium having stored thereon computer-executable instructions, which when executed by a processor, perform the method steps of any one of claims 1-7.