CN106054858B - The method of the vehicle remote diagnosis and spare part retrieval classified based on decision tree classification and error code - Google Patents
The method of the vehicle remote diagnosis and spare part retrieval classified based on decision tree classification and error code Download PDFInfo
- Publication number
- CN106054858B CN106054858B CN201610365652.4A CN201610365652A CN106054858B CN 106054858 B CN106054858 B CN 106054858B CN 201610365652 A CN201610365652 A CN 201610365652A CN 106054858 B CN106054858 B CN 106054858B
- Authority
- CN
- China
- Prior art keywords
- error code
- spare part
- word
- code
- cell
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0208—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the configuration of the monitoring system
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The method of the vehicle remote diagnosis and spare part retrieval classified based on decision tree classification and error code, belong to information retrieval field, since different spare parts is in the identical situation of main group number, packet number is still variant, in order to solve the problems, such as accurately to match spare part, has technical point that and error code is identified and is classified；Parsing vehicle VIN code obtains variable, and the variable includes parsing obtained engine displacement, body style, engine mission type by VIN code；Decision tree analysis is done to spare part code corresponding to variable, variable data is completed and sorts out to form spare part information, and establish index, form diagnostic knowledge base；Language model is created, cell dictionary is established, word cutting retrieves cell word in the cell dictionary, and arranges the cell word, using the Decision Classfication of decision-tree model, forms the diagnostic data base that error code corresponds to work item；Diagnostic data base is associated with diagnostic knowledge base, and establishes major key；Effect is: after obtaining error code, can be quickly found out the solution of most common failure and correspond to spare part, work item.
Description
Technical field
The invention belongs to information retrieval fields, are related to a kind of method for vehicle remote diagnosis and spare part retrieval
Background technique
The rank that China's automobile maintenance industry has been diagnosed from the feeling and practical experience that fully rely on examiner at present
Section has developed to and has carried out comprehensive detection diagnostic phases using special equipment, but generally existing in orthodox car maintenance industry
Many problems, such as service worker's technology aging, often can not quickly, economically be solved using the technical force of various aspects therefore
Barrier；Increasing with car ownership, market respective services emerge in multitude like the mushrooms after rain after automobile.So from car owner
How angle more preferably could more fully understand vehicle condition, when breaking down, how quick obtaining love vehicle scheme to be solved and required
Working hour and spare part relevant information, accurately automobile wearable device is completely necessary to car owner's real-time requirement is met.It is general
OBD mobile unit can only read associated vehicle fault message, and detailed settlement project and related maintenance people cannot be made to failure
Expenses of labour, spare part take, so that car owner be caused blindly into shop, blindly to consume.
Summary of the invention
When breaking down to solve host vehicle, work Xiang Yubei corresponding to the error code can be accurately and quickly matched
Part, the following technical solutions are proposed by the present invention: a kind of vehicle remote diagnosis and spare part classified based on decision tree classification and error code
The method of retrieval, including
Step 1 acquires vehicle information data；
Step 2 is identified and is classified to error code；
Step 3 parsing vehicle VIN code obtains variable, the variable include parsed by VIN code obtained engine displacement,
Body style, engine mission type；
Step 4 does decision tree analysis to spare part code corresponding to variable, completes variable data and sorts out to form spare part letter
Breath, and index is established, form diagnostic knowledge base；
Step 5 creates language model, establishes cell dictionary, and word cutting retrieves cell word in the cell dictionary, side by side
The cell word is arranged, using the Decision Classfication of decision-tree model, forms the diagnostic data base that error code corresponds to work item；
Diagnostic data base is associated with by step 6 with diagnostic knowledge base, and establishes major key；
Step 7 identifies the error code that vehicle trouble generates, and parses the variable that vehicle VIN code obtains by keyword
To carry out systematic searching, work item and spare part information are obtained.
Beneficial aspects: the present invention after obtaining error code, can be quickly found out the solution of most common failure and correspond to standby
Part, work item.The experience confinement problems for effectively solving technician and partsman obtain the solution of failure from big data.
It has the step of identifying and classifying to error code to the present invention, can judge when difference occurs in the error code between different automobile types
Its consistency；And step five of the invention constructs one and decomposes from failure-description to term vector, establishes maintenance cell dictionary；
Word cutting by failure-description in cell dictionary is retrieved, and according to the syntactic rule that cell word weight is established, is realized from decision process
The maintenance decision tree for arriving maintenance solution again to decision recommendation, finally realizes Remote Fault Diagnosis.
Detailed description of the invention
Fig. 1 is flow chart of the present invention for vehicle remote diagnosis and the method for spare part retrieval；
Fig. 2 is the VIN number code translation schematic diagram for the vehicle that chassis number is LFV5A14B8Y3000001.
Specific embodiment
Embodiment 1: a method of the vehicle remote diagnosis and spare part classified based on decision tree classification and error code are retrieved,
Including
Step 1 acquires vehicle information data；
Step 2 is identified and is classified to error code；
Step 3 parsing vehicle VIN code obtains variable, the variable include parsed by VIN code obtained engine displacement,
Body style, engine mission type；
Step 4 does decision tree analysis to spare part code corresponding to variable, completes variable data and sorts out to form spare part letter
Breath, and index is established, form diagnostic knowledge base；
Step 5 creates language model, establishes cell dictionary, and word cutting retrieves cell word in the cell dictionary, side by side
The cell word is arranged, using the Decision Classfication of decision-tree model, forms the diagnostic data base that error code corresponds to work item；
Diagnostic data base is associated with by step 6 with diagnostic knowledge base, and establishes major key；
The error code that step 7 generates vehicle trouble identifies, and by keyword progress systematic searching, obtain work item and
Spare part information.
Embodiment 2:With technical solution same as Example 1, more specifically, four the step of for embodiment 1
It says,
Using the historical record of maintenance and repair parts table as data basis in the step 4, spare part is divided by decision-tree model
Class, maintenance and repair parts table sample is for example shown in table one:
Table one
| VIN123 | VIN4 | VIN6 | VIN78 | BJDM |
| LFV | 5 | 1 | 4B | 06J115403J |
| LFV | 3 | 2 | 8K | LN052167A21 |
| LFV | 4 | 2 | 4F | LN052167A24 |
The basic principle of decision-tree model is as follows:
First: determining the entropy of every dimension spare part different classifications, by taking VIN4 as an example, entropy is defined as
E=sum (- p (I) * log (p (I)))
Wherein I=1:N (N class as a result, such as a kind of this example, i.e. the spare part belongs to this vehicle, therefore probability P (I)=1)
Then E (5)=- (1/1) Log2 (1/1)-(0/1) Log2 (0/1)=0+0=0
E (3)=- (1/1) Log2 (1/1)-(0/1) Log2 (0/1)=0+0=0
E (4)=- (1/1) Log2 (1/1)-(0/1) Log2 (0/1)=0+0=0
If entropy is 0, show that discrimination is higher；Entropy is 1, then shows no discrimination；
Therefore these three different spare part codes can be distinguished by VIN4.
After having determined how each dimension classifies, the priority level between different dimensions distinguishes Gain by information gain
(Sample, Action)=E (sample)-sum (| Sample (v) |/Sample*E (Sample (v))) then Gain (VIN4)=
E (S)-(1/3) * E (5)-(1/3) * E (3)-(1/3) * E (4)=1-0=1Gain (VIN6)=E (S)-(1/3) * E (1)-(2/
3) * E (2)=1-0-2/3=1/3Gain (VIN78)=E (S)-(1/3) * E (4B)-(1/3) * E (8K)-(1/3) * E (4F)=
1-0=1
If information gain is bigger, show that classification priority is higher；Conversely, priority is lower.
So chassis number the 4th (VIN4) is identical with chassis number 78 (VIN78) classification priority, followed by chassis
Number the 6th (VIN6).
It, can be by spare part code according to chassis number the 4th (VIN4), chassis number the 6th (VIN6) by the above committed step
It is distinguished with chassis number 78 (VIN78).
In conclusion the basic step of spare part search method is:
Maintenance and repair parts table is distinguished with dimension by comentropy；
Maintenance and repair parts table different dimensions are divided into priority by information gain；
The priority and differentiation degree divided according to 1,2 steps draws decision tree；
A regular chassis number is inputted, system is defeated according to VIN123, VIN4, VIN6, VIN78 and the decision tree obtained
Spare part code under the vehicle out.
The spare part code by association price of spare parts table, obtain Chinese, price and the present use state of spare part with
And the applicable vehicle model information of spare part.
The acquisition of above-mentioned technical proposal, be will different automobile types, different displacements, corresponding to different engine mission type
Spare part code sort out carry out analysis comparison one by one after, find under the premise of identical main group number, the information such as vehicle discharge capacity are different, institute
Corresponding spare part code is also not quite similar, wherein regular in order to look for, and has used the above method, more complete and comprehensive to be formed
Theoretical knowledge information bank.
Embodiment 3:With technical solution identical with embodiment 1 or 2, more specifically, five the step of for embodiment 1
For,
Creation language model in the step 5 is established cell dictionary and is comprised the following steps:
S1.1 acquires professional failure-description language；
S1.2 carries out term vector decomposition to the professional failure-description language.
The creation of the language model appearance based on n-th of cell word and the front n-1 cell word phases
The hypothesis of pass；There is the calculation formula of weight in the one failure-description sentence T are as follows:
P (T)=P (w1,w2,w3,…,wn)
=P (w1)×P(w2|w1)×P(w3|w1,w2)×…×P(wn|w1,w2,…,wn-1)
≈P(w1)×P(w2|w1)×P(w3|w2)…P(wn|wn-1)；
Wherein, P (T) is the weight of the failure-description sentence T, P (wn|w1,w2,…,wn-1) it is n-th of cell word
Weight.
Word cutting retrieves cell word in cell dictionary in the step 5, and the step of arranging the cell word is:
S2.1 carries out word cutting retrieval for the failure-description sentence T in the cell dictionary；
If S2.2 retrieves the cell word, the weight of the cell word is calculated；
Cell word described in big minispread of the S2.3 according to the cell word weight.
If S2.4 does not retrieve the cell word, the cell word that this is not retrieved is stored in newly-increased cell word
Library.
The calculation method of cell word weight described in the S2.2 step are as follows:
S2.2.1 calculates the chi-square statistics amount of each cell word；
S2.2.2 takes highest i-th of cell word of the chi-square statistics amount score value, calculates i-th of cell word in jth
The number occurred in a failure-description；The calculation method of chi-square statistics amount score value described in the S2.2.1 step are as follows:
Weight=round ((10 × (1+lg (tfij))/(1+lg(lj))),n)；
Wherein tfijFor the number that i-th of cell word occurs in j-th of failure-description, ljFor the length of j-th of failure-description
Degree, n are the integer of 3-6.
Using the Decision Classfication of decision-tree model, the diagnostic data base that error code corresponds to work item is formed, in the step, also wrap
It includes and component measured value is combined to provide maintenance solution (work item).
More specifically,
The present embodiment converts Machine self-learning for the problem of natural language understanding by reaching natural language digitlization
The problem of.
(1) basic corpus is constructed
The method for creating basic corpus is to give a character string, the probability of its natural language be P (w1, w2,
W3 ..., wn), w1 to wn is successively each cell word of the words.
Model based on it is such a it is assumed that n-th of word appearance only it is related to the word of front n-1, and with other any words
All uncorrelated, the probability of whole sentence is exactly the product of each participle probability of occurrence.
For a failure-description T, its probability calculation formula are as follows:
P (T)=P (w1,w2,w3,…,wn)
=P (w1)×P(w2|w1)×P(w3|w1,w2)×…×P(wn|w1,w2,…,wn-1)
≈P(w1)×P(w2|w1)×P(w3|w2)…P(wn|wn-1)
(2) term vector decomposes: term vector is the sentence being cut into after several cell words.Each failure-description T can table
It is shown as the term vector of n dimension, T=(w1,w2,...,wn), wherein w1To wnFor each cell word for including in each term vector.
By taking vehicle failure as an example:
Failure-description are as follows:
P0021 A camshaft location (bank of cylinder 2)-timing is excessively advanced
Indicate that above-mentioned failure-description sentence, above-mentioned failure-description sentence can use term vector (w with term vector1,w2,w3,…,
wn).Wherein, w1For camshaft, w2For position, w3For cylinder, w4For timing, w5It is excessive.Each cell word is chosen as feature
, wherein wiIndicate ith feature item.
Then the chi-square statistics amount for calculating each cell word takes highest i-th of cell word conduct of score value in training set
Word is represented, to find out and term vector (w1,w2,w3,…,wn) the smallest cell word of correlation.
wiWeight calculation formula are as follows:
Weight=round ((10 × (1+lg (tfij))/(1+lg(lj))),3)；
Wherein tfijFor the number that i-th of term vector occurs in j-th of failure-description, ljFor the length of j-th of failure-description
Degree.Round function is rounded up to specified decimal place, is third position after decimal point in above formula.
Calculate the feature statement of the failure-description term vector after weight are as follows:
| Vector participle | Weight |
| Camshaft | 0.286 |
| Position | 0.191 |
| Cylinder | 0.048 |
| Timing | 0.048 |
| Excessively | 0.048 |
(1) arranging cells word is retrieved:
As shown in Figure 1, carrying out term vector after user's input fault describes T to failure-description T first and decomposing complete language
Sentence is cut into multiple cell contaminations.Then it is directed to each cell word, word cutting retrieval is carried out in cell dictionary.If retrieval
It arrives, then calculates the weight of the cell word；If do not retrieved, the cell word that this is not retrieved is stored in newly-increased industry
Cell dictionary, then retrieved for remaining cell word, calculate its probability.
According to each cell word retrieved in cell dictionary and weight is calculated, according to weight arranging cells from big to small
Word.
(2) result is analyzed:
By taking vehicle failure as an example:
Failure-description are as follows:
P0021 A camshaft location (bank of cylinder 2)-timing is excessively advanced
Each cell word for retrieving in cell dictionary simultaneously calculates weight, according to weight arranging cells word from big to small:
| Characteristic item | Weight |
| Camshaft | 0.286 |
| Position | 0.191 |
| Actuator | 0.095 |
| Sensor | 0.048 |
| Conversion | 0.048 |
| Response | 0.048 |
| Excessively | 0.048 |
| Circuit | 0.048 |
| Cylinder | 0.048 |
| It adjusts | 0.048 |
| Exhaust | 0.048 |
| Timing | 0.048 |
(3) techniques of discriminant analysis is used:
The term vector sequence of failure-description T is T=(w1, w2, w3 ..., wn), and each cell word calculates in failure-description T
Sequence after weight is S=(s1, s2, s3 ..., sn).The calculation formula of term vector classification and matching maximum probability are as follows: W=
argmaxP(S|T)。
It is classified according to the syntactic rule that term vector classification and matching maximum probability arranges from high to low:
| Classification | Rule |
| Level-one | Camshaft |
| Second level | Position |
| Three-level | Timing |
(4) decision tree analysis
Result according to above-mentioned syntactic rule for the classification of cell word establishes decision tree.
Disaggregated model 1
As shown in Fig. 2, camshaft is level-one, position converts, is adjusted to second level, sensor, actuator, response, timing,
Exhaust is three-level.Corresponding solution is respectively: sensor corresponds to CMPS Camshaft Position Sensor, engine wiring harness and starts
Machine control unit, actuator correspond to engine wiring harness, control unit of engine and camshaft adjuster, and timing corresponds to engine line
Beam, control unit of engine, camshaft adjuster, camshaft stretcher and timing chain, and so on.The solution that block arrow is directed toward
Certainly scheme indicates to need replacing the probability of this accessory much larger than the probability for replacing the accessory that other thin arrows are directed toward, this probability is root
It is preset in decision-tree model according to a large amount of mantenance datas.Therefore, by failure-description " camshaft location (cylinder
Column 2)-timing is excessively advanced " carry out term vector decomposition and retrieved in cell dictionary, obtain camshaft, position and timing
Three cell words, be computed cell word weight and according to syntactic rule to cell word carry out arrangement show that camshaft is located at level-one,
Position is located at second level, and timing is located at three-level.By the Decision Classfication of decision-tree model, most possible maintenance solution is obtained
To replace or repairing camshaft stretcher.
Disaggregated model 2
To failure-description sentence " CMPS Camshaft Position Sensor=insincere signal of > sensor " carry out term vector decomposition and
It is retrieved in cell dictionary, obtains three camshaft, position, sensor cell words.Cell word weight is calculated for three cell words
And arrangement is carried out to cell word according to syntactic rule and show that camshaft is located at level-one, position is located at second level, and sensor is located at three-level.
In conjunction with component measured value, the accessory needed replacing is judged.First checking for wire and plug, whether there is or not poor contact, syringe needle bendings.
If it is, needing replacing wire and plug.If it is not, then checking the supply of camshaft-signal sensor voltage into next stage decision tree
Whether between 4.5-5.5V.If it is the signal voltage of control unit of engine is checked, if between 4.5-5.5V.Such as
Otherwise fruit needs replacing engine wiring harness.And so on, by the Decision Classfication combination component measured value of decision-tree model, give
Solution is repaired out.
The present embodiment can remotely and rapidly identify the failure of equipment, accurately provide effectively maintenance solution.Inspection
Mark standard is unified, avoids due to artificial judgment bring error.And save a large amount of manpower and material resources.
Example IV:With with embodiment 1 or 2 or 3 identical technical solutions, more specifically, for the step of embodiment 1
For rapid two,
The method that error code is identified and is classified, comprising:
Naive Bayes Classifier, specific steps are constructed first with training dataset are as follows:
S1: manual sort is carried out to the error code in training set, the classification after manual sort integrates as c:
C={ c1, c2..., ci..., cn}；
S2: the error code in each classification is segmented to and is calculated the TF-IDF weight value of each word；
TFiIt is characterized the word frequency of word i；wiIt is characterized the number that word i occurs in all error code, ∑jwjIndicate error code
In all Feature Words frequency of occurrence summation；
IDFiIt is characterized the reverse document-frequency of word i；E is the sum of error code in corpus, { k:wi∈ekIndicate include
The failure yardage of Feature Words i；
TF-IDF weight value are as follows: TF-IDF=TF × IDF, that is, be the product of word frequency Yu reverse document-frequency；
S3: filtering out Feature Words according to TF-IDF weight value and establishes feature dictionary, and error code e is defined as several spies
Levy the set of word:
E={ w1, w2..., wn}
S4: calculating separately conditional probability of the Feature Words under the conditions of category in each classification, according to Bayes' theorem meter
It calculates some error code e and belongs to ciThe probability of class failure, specific formula is as follows:
P(ci| e)=[P (e | ci)P(ci)]/P(e)
Wherein, P (e) is the probability that an error code e is randomly selected from failure code space；P(ci) it is an error code
Classification ciThe ratio shared by failure code space；P(e|ci) it is for given error code classification ciThe probability of occurrence of middle error code e；
For each error code classification, the probability that an error code is randomly selected from failure code space is just as
, therefore P (e) can ignore and not calculate, such formula (1) can be written as follow form:
P(ci|e)∝P(e|ci)P(ci)
Some error code probability in each classification is calculated, taking the classification for possessing most probable value is the classification of the error code
As a result, specifically:
G(e)≡argmax{P(e|ci)P(ci)}。
S5: building Naive Bayes Classifier.
Then specifically classified to new error code, specifically:
More specifically, specifically classified to new error code, the error code being located in classification i is ei, while there are also to
Matched error code enew, their participle situation is as follows:
ei={ w1, w2, w3, w4}
enew={ w2, w3, w5}
Step 1: a vector space is constituted by this two groups feature set of words:
Step 2: in conjunction with vector space, respectively obtain the term vector value of two error code:
Step 3: two error code approximation situations are calculated using cosine similarity, specific formula is as follows:
Step 4: if similarity value calculated is greater than 80%, may determine that the two error code are identical；Such as
Fruit similarity value calculated is less than threshold value, then continues similar to other classification error code progress cosine in known fault code library
Degree calculates.
Using the above technical method, following technical effect can be obtained: the failure for different automobile types of accurately classifying automatically
Number is restrained with hundred million grades of error code, realizes the exploration of failure problems, solution between different depot's vehicles by code.It can section
A large amount of labour is saved, shortens the duty cycle, reduces cost.
The preferable specific embodiment of the above, only the invention, but the protection scope of the invention is not
It is confined to this, anyone skilled in the art is in the technical scope that the invention discloses, according to the present invention
The technical solution of creation and its inventive concept are subject to equivalent substitution or change, should all cover the invention protection scope it
It is interior.
Claims (9)
1. a kind of method of the vehicle remote diagnosis classified based on decision tree classification and error code and spare part retrieval, feature are existed
In, including
Step 1 acquires vehicle information data；
Step 2 is identified and is classified to error code；
Step 3 parsing vehicle VIN code obtains variable, and the variable includes the engine displacement obtained by the parsing of VIN code, vehicle body
Type, engine mission type；
Step 4 does decision tree analysis to spare part code corresponding to variable, completes variable data and sorts out to form spare part information, and
Index is established, diagnostic knowledge base is formed；
Step 5 creates language model, establishes cell dictionary, and word cutting retrieves cell word in the cell dictionary, and arranges institute
Cell word is stated, using the Decision Classfication of decision-tree model, forms the diagnostic data base that error code corresponds to work item；
Diagnostic data base is associated with by step 6 with diagnostic knowledge base, and establishes major key；
The error code that step 7 generates vehicle trouble identifies, and by variable that keyword parsing vehicle VIN code obtains with into
Row systematic searching obtains work item and spare part information.
2. the side as described in claim 1 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, which is characterized in that using the historical record of maintenance and repair parts table as data basis in the step 4, by decision-tree model to standby
Part, which is done, classifies；
The step of spare part search method, is:
(1) maintenance and repair parts table is distinguished with dimension by comentropy；
(2) maintenance and repair parts table different dimensions are divided into priority by information gain；
(3) priority and differentiation degree divided according to step (1), (2) draws decision tree；
(4) a regular VIN code is inputted, according to the spare part code under VIN code and the decision tree obtained output vehicle.
3. the side as claimed in claim 2 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, which is characterized in that the price list of the spare part associated codes spare part, the spare part information include the Chinese of spare part, valence
The applicable vehicle model information of lattice and present use state and spare part.
4. the side as described in claim 1 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, which is characterized in that the creation language model in the step 5 is established cell dictionary and comprised the following steps:
S1.1 acquires professional failure-description language；
S1.2 carries out term vector decomposition to the professional failure-description language.
5. the side as claimed in claim 4 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, it is characterised in that: appearance of the creation of the language model based on n-th of cell word is only described thin with front n-1
The relevant hypothesis of born of the same parents' word；There is the calculation formula of weight in one failure-description sentence T are as follows:
P (T)=P (w1,w2,w3,…,wn)
=P (w1)×P(w2|w1)×P(w3|w1,w2)×…×P(wn|w1,w2,…,wn-1)
≈P(w1)×P(w2|w1)×P(w3|w2)…P(wn|wn-1)；
Wherein, P (T) is the weight of failure-description sentence T, P (wn|w1,w2,…,wn-1) be n-th of cell word weight.
6. the side as described in claim 1 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, which is characterized in that word cutting retrieves cell word, and the step of arranging the cell word in the cell dictionary in the step 5
It is:
S2.1 carries out word cutting retrieval for a failure-description sentence T in the cell dictionary；
If S2.2 retrieves the cell word, the weight of the cell word is calculated；
Cell word described in big minispread of the S2.3 according to the cell word weight.
7. the side as described in claim 1 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, which is characterized in that the method that error code is identified and is classified, comprising:
Naive Bayes Classifier is constructed first with training dataset；
Then specifically classified to new error code；
The step of constructing Naive Bayes Classifier are as follows:
S1: manual sort is carried out to the error code in training set；
S2: the error code in each classification is segmented to and is calculated the TF-IDF weight value of each word；
S3: filtering out Feature Words according to TF-IDF weight value and establishes feature dictionary；
S4: conditional probability of the Feature Words under the conditions of category in each classification is calculated separately；
S5: building Naive Bayes Classifier.
8. the side as claimed in claim 7 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, which is characterized in that in step S2TFiIt is characterized the word frequency of word i；wiIt is faulty in institute to be characterized word i
The number occurred in code, ∑jwjIndicate the frequency of occurrence summation of all Feature Words in error code；
In step S2IDFiIt is characterized the reverse document-frequency of word i；
E is the sum of error code in corpus, { k:wi∈ekIndicate the failure yardage comprising Feature Words i；
TF-IDF weight value in step S2 are as follows:
TF-IDF=TF × IDF
TF-IDF weight value is the product of word frequency Yu reverse document-frequency.
9. the side as claimed in claim 7 based on decision tree classification and the error code vehicle remote diagnosis classified and spare part retrieval
Method, which is characterized in that specifically classified to new error code, the error code being located in classification i is ei, while there are also a to be matched
Error code enew:
Step 1: a vector space is constituted by this two groups feature set of words:
Step 2: in conjunction with vector space, respectively obtain the term vector value of two error code:
Step 3: calculating two error code approximation situations using cosine similarity；
Step 4: if similarity value calculated is greater than threshold value, may determine that the two error code are identical；If institute
The similarity value of calculating is less than threshold value, then continues to carry out cosine similarity meter with other classification error code in known fault code library
It calculates；Two error code approximation situations are judged in step 3, specific formula is as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610365652.4A CN106054858B (en) | 2016-05-27 | 2016-05-27 | The method of the vehicle remote diagnosis and spare part retrieval classified based on decision tree classification and error code |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610365652.4A CN106054858B (en) | 2016-05-27 | 2016-05-27 | The method of the vehicle remote diagnosis and spare part retrieval classified based on decision tree classification and error code |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106054858A CN106054858A (en) | 2016-10-26 |
| CN106054858B true CN106054858B (en) | 2019-09-27 |
Family
ID=57174983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610365652.4A Active CN106054858B (en) | 2016-05-27 | 2016-05-27 | The method of the vehicle remote diagnosis and spare part retrieval classified based on decision tree classification and error code |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106054858B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108573198A (en) * | 2017-03-14 | 2018-09-25 | 优信互联（北京）信息技术有限公司 | A kind of method and device identifying vehicle information according to Vehicle Identify Number |
| CN107193929B (en) * | 2017-05-17 | 2020-12-25 | 田兆杰 | Vehicle fault question-answering method and device based on feature extraction and similarity calculation |
| CN108985465A (en) * | 2018-05-21 | 2018-12-11 | 许继电气股份有限公司 | A kind of converter station Fault Classification and system |
| CN109032107B (en) * | 2018-06-05 | 2021-07-20 | 国家电网公司 | Equipment fault signal frequency prediction method based on Bayesian classification |
| US11567487B2 (en) * | 2018-09-28 | 2023-01-31 | Abb Schweiz Ag | Fault diagnosis system and method for electric drives |
| CN110413490B (en) * | 2019-08-01 | 2023-07-14 | 阿波罗智能技术(北京)有限公司 | Method for determining error information type, error information code classification method and device |
| CN112943913B (en) * | 2021-02-26 | 2022-05-20 | 北京理工大学 | AMT (automated mechanical transmission) gear engaging process redundancy control method based on decision tree algorithm |
| CN113469833B (en) * | 2021-06-17 | 2023-08-29 | 南京大学 | Remote intelligent detection and processing method supporting industrial Internet equipment |
| CN114647234B (en) * | 2022-05-23 | 2022-08-09 | 医链数科医疗科技(江苏)有限公司 | Medical equipment monitoring method and device based on Internet of things and storage medium |
| CN117556829B (en) * | 2024-01-11 | 2024-04-09 | 中汽信息科技(天津)有限公司 | Comparison method, device and equipment of vehicle fault codes and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102496028A (en) * | 2011-11-14 | 2012-06-13 | 华中科技大学 | Breakdown maintenance and fault analysis method for complicated equipment |
| CN102521613A (en) * | 2011-12-17 | 2012-06-27 | 山东省科学院自动化研究所 | Method for fault diagnosis of automobile electronic system |
| CN103001811A (en) * | 2012-12-31 | 2013-03-27 | 北京启明星辰信息技术股份有限公司 | Method and device for fault locating |
| CN104392072A (en) * | 2014-12-12 | 2015-03-04 | 北京交通大学 | Equipment integration system reliability analysis method based on dynamic fault tree |
-
2016
- 2016-05-27 CN CN201610365652.4A patent/CN106054858B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102496028A (en) * | 2011-11-14 | 2012-06-13 | 华中科技大学 | Breakdown maintenance and fault analysis method for complicated equipment |
| CN102521613A (en) * | 2011-12-17 | 2012-06-27 | 山东省科学院自动化研究所 | Method for fault diagnosis of automobile electronic system |
| CN103001811A (en) * | 2012-12-31 | 2013-03-27 | 北京启明星辰信息技术股份有限公司 | Method and device for fault locating |
| CN104392072A (en) * | 2014-12-12 | 2015-03-04 | 北京交通大学 | Equipment integration system reliability analysis method based on dynamic fault tree |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106054858A (en) | 2016-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106054858B (en) | The method of the vehicle remote diagnosis and spare part retrieval classified based on decision tree classification and error code | |
| CN106021545B (en) | Method for vehicle remote diagnosis and spare part retrieval | |
| CN104809108B (en) | Information monitoring analysis system | |
| CN106095785B (en) | Fault code diagnosis vehicle work item and spare part retrieval method based on decision tree classification | |
| CN106056221A (en) | FP-Tree sequence pattern mining and fault code classification-based vehicle remote diagnosis and spare part retrieval method | |
| CN114048870A (en) | Power system abnormity monitoring method based on log characteristic intelligent mining | |
| US20210073216A1 (en) | Business intelligence system based on artificial intelligence and analysis method thereof | |
| CN110222192A (en) | Corpus method for building up and device | |
| CN111581956A (en) | Sensitive information identification method and system based on BERT model and K nearest neighbor | |
| CN115618014B (en) | Standard document analysis management system and method applying big data technology | |
| CN113313409A (en) | Power system secondary equipment defect analysis method and system based on data association | |
| CN115544272A (en) | Attention mechanism-based chemical accident cause knowledge graph construction method | |
| CN103544307A (en) | Multi-search-engine automatic comparison and evaluation method independent of document library | |
| CN113065356A (en) | IT equipment operation and maintenance fault suggestion processing method based on semantic analysis algorithm | |
| CN114048318A (en) | Clustering method, system, device and storage medium based on density radius | |
| CN113435759A (en) | Primary equipment risk intelligent evaluation method based on deep learning | |
| CN117290404A (en) | Method and system for rapidly searching and practical main distribution network fault processing method | |
| CN111783456A (en) | Public opinion analysis method using semantic analysis technology | |
| CN109902173B (en) | Chinese text classification method | |
| CN116860977B (en) | Abnormality detection system and method for contradiction dispute mediation | |
| CN106054857B (en) | Maintenance decision tree/word vector-based fault remote diagnosis platform | |
| CN104809253B (en) | Internet data analysis system | |
| CN115310869B (en) | Combined supervision method, system, equipment and storage medium for supervision items | |
| CN106056223B (en) | Platform for vehicle remote diagnosis and spare part retrieval | |
| CN106055439B (en) | Based on maintenance decision tree/term vector Remote Fault Diagnosis system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |