CN112380206A - Diagnosis and repair method of traffic time sequence data - Google Patents
Diagnosis and repair method of traffic time sequence data Download PDFInfo
- Publication number
- CN112380206A CN112380206A CN202011338594.9A CN202011338594A CN112380206A CN 112380206 A CN112380206 A CN 112380206A CN 202011338594 A CN202011338594 A CN 202011338594A CN 112380206 A CN112380206 A CN 112380206A
- Authority
- CN
- China
- Prior art keywords
- data
- repair
- current time
- current
- alpha
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a method for diagnosing and repairing traffic time sequence data, which comprises the following steps: s1: starting a data preprocessing algorithm to collect data; s2: analyzing and comparing historical data with current data to realize data diagnosis; s3: and performing data repair on the current data by using the historical data. The invention provides a method for diagnosing and repairing traffic time sequence data, which can process traffic time sequence data of various different sources and types, diagnose the data, repair the data which is judged to be missing or has higher abnormal level after diagnosis, analyze the confidence degree of the repaired data and finally output more reliable data with marks.
Description
Technical Field
The invention relates to the technical field of traffic control, in particular to a diagnosis and repair method of traffic time sequence data.
Background
At present, it is a common consensus in the industry to perform certain preprocessing on traffic time series data, but none of the prior arts can perform diagnostic marking on data with poor correlation from multiple different sources by using a statistical means and perform repairing by using time and space correlation between the same traffic data, and none of the prior arts can perform general preprocessing on most traffic time series data.
The existing methods for diagnosing traffic time series data mainly comprise a threshold value method, a curve fitting method, an interval estimation method and the like. These methods estimate the reasonable range of data based on different theories: the threshold method is mainly characterized in that a threshold upper limit and a threshold lower limit are set according to practical experience, and data which obviously do not accord with practical conditions are removed; the curve fitting method is that a fitting curve is drawn according to the relation between all traffic data of a traffic flow theory, and then reasonable data are screened according to the deviation degree of actual data and the fitting curve; the interval estimation method is to determine a central value, which is usually a mean value or an expected value of the data, and then to retain data within a certain range of the central value, and to consider the remaining data beyond the range as error data.
The existing methods for repairing traffic time series data mainly comprise a substitution method, a moving average method, a model prediction method and the like. The replacement method is to replace missing data or error data with data such as an average number, a median number or a mode number according to the characteristics of the data for different data; the moving average method adopts a moving average model, and smoothes the transverse time sequence of the traffic data by using the weighted average of historical contemporaneous data, so that the noise caused by various uncertain factors can be reduced; the model prediction method is to train a complete historical data set by machine learning methods such as a generative confrontation network model, a space-time learning network, a random forest algorithm and the like, and then predict a data set where missing data or noisy data is located so as to replace the missing or wrong data. However, these methods do not provide a complete diagnostic, labeling, and repair procedure that is applicable to most traffic time series data.
The invention discloses a method for diagnosing and repairing the quality of road network traffic time sequence characteristic data, which is disclosed by Chinese patent publication No. CN111179591A, No. 2020, 05 and 19, and comprises the following steps: s1, obtaining a regional road network topological structure, and constructing a traffic data prediction model; s2, acquiring historical traffic data, training a model and checking the prediction precision of the model; s3, fusing the traffic data acquired by the traffic detector with the traffic data generated, and detecting and repairing the abnormal data in real time; and S4, performing incremental training on the prediction model to ensure the prediction precision of the model. According to the invention, the regional road network topological structure does not need to be acquired and the traffic data prediction model does not need to be constructed, so that the model does not need to be trained, the incremental training is also not needed, the complexity of the system is reduced, the maintainability of the system is increased, and the contrast patent does not carry out confidence judgment on the prediction result, so that feedback adjustment cannot be carried out, and a better prediction effect is obtained.
Disclosure of Invention
The invention aims to solve the problem that the prior art can not perform feedback judgment on traffic time sequence repair data, and provides a diagnosis and repair method of traffic time sequence data, which can process traffic time sequence data of various sources and types, diagnose the data, repair the data which is judged to be missing or have higher abnormal grade after diagnosis, analyze the confidence degree of the repaired data, and finally output the marked and more reliable data.
In order to achieve the purpose, the invention adopts the following technical scheme:
the technical scheme adopted by the invention for solving the technical problems is as follows: a method for diagnosing and repairing traffic time series data comprises the following steps:
s1: starting a data preprocessing algorithm to collect data;
s2: analyzing and comparing historical data with current data to realize data diagnosis;
s3: and performing data repair on the current data by using the historical data.
Preferably, the step S1 includes the following steps: and after the data preprocessing algorithm is started, starting to import data from each data import component. After the externally input traffic time sequence data is taken, a series of operations such as diagnosis, restoration and marking are carried out on the original data, more complete and accurate data are finally output, and the application effect of the data is improved.
Preferably, the step S2 includes the steps of:
s21: acquiring data x corresponding to the current moment;
s22: inquiring and acquiring current-time forward data and current-time historical data of the input data in the S1, wherein the acquisition quantity of the forward data and the historical data is configurable; diagnosing the current time data according to the current time forward data and the current time historical data;
s23: taking a union set of current-time forward data and current-time historical data, and calculating an upper quartile Q1, a lower quartile Q3 and a quartile interval IQR (Q3-Q1) of the union set, wherein if the current-time data x is in a range of (Q3-1.5IQR, Q1+1.5IQR), a data abnormality level p (normal); otherwise, entering the calculation of the abnormal degree. If 1.5IQR < Q3-x < alpha2Iqr or 1.5IQR < x-Q1 < alpha1IQR, then p is slot; if α is1IQR＜Q3-x＜α2Iqr or alpha1IQR＜x-Q1＜α2IQR, then p ═ middle; otherwise p is severe. Wherein alpha is1，α2Are all configurable parameters. .
Preferably, in step S21, if there is no current time data, the process proceeds to S3 repair flow as it is with a flag p ═ lose.
Preferably, the step S3 includes the steps of:
s31: repairing data with abnormal grade p ═ segment or p ═ lose;
s32: sequencing the acquired forward data at the current time and the historical data at the current time from small to large, taking the first j data (j is a set parameter) with the smallest distance, taking the weighted average of the data as repair data, and taking the weight as the reciprocal of the distance between the data and the data at the current time:
wherein xiFor obtaining current time forward data or current time historical data, LiIs corresponding to xiAnd the distance between the current time data;
s33: and setting confidence coefficient, and prompting the user data by using the confidence coefficient. .
Preferably, in step S31, the process of repairing the data with the abnormal level p-segment or p-lose is as follows: inquiring and acquiring data with the abnormal degree lower than midle in the current time forward data and the current time historical data of the input data in the S1; if certain calendar history data is missing, the following steps are executed:
a) data before and after the history data at the time Δ t exist: the average value of Δ t moments before and after the history data is used for replacing:
xnfor missing historical data, xn-1History data at a time Δ t immediately before the missing history data, xn+1History data at a delta t moment after the history data is lost; recording the distance L between the substitute value and the current time datarIs the original distance LoSquare of sum of squares with Δ t/2:
b) data loss at Δ t before and after the history data: replacing data at a delta t moment before the current moment; recording the distance L between the substitute value and the current time datarIs the original distance LoAnd the square sum of Δ t to the square:
can process traffic time sequence data of various sources and types, diagnose the data, and judge the diagnosed data as missing orAnd repairing the data with higher abnormal grade, analyzing the confidence degree of the repaired data, and finally outputting the marked and more reliable data.
Preferably, the process of setting the confidence level in step S33 is as follows: noting the degree of confidence alpha0High. Calculating the number k of the acquired forward data at the current time and the current time historical data, and calculating the maximum distance L between the acquired data and the current time data; if b is not more than k and is less than c, then alpha0If L > L, then alpha is determined0Or else α0Midle; if a is not more than k and is less than b, then alpha0If L > L, then alpha is determined0False, otherwise α0Low; if 0 < k < a, then alpha0False; if k is 0, then α0Fail; otherwise alpha0High; wherein a, b, c and l are all set parameters. c is a first threshold of the repair number, and when the number of the data used for repair is larger than or equal to c, the data used for repair is considered to be enough in quantity, and the repair is reliable; when the number of the data used for repairing is less than c, the data used for repairing is considered to be insufficient, and the confidence coefficient of the repaired data is reduced; b is a second threshold of the repair number, and when the number of the data used for repair is less than b, the data used for repair is considered to be less in quantity, and the confidence of the repair data is continuously reduced; a is a third threshold of the repair number, and when the number used for repair is smaller than a, the repair is unreliable due to the fact that the data volume used for repair is too small; and l is a repair length threshold, and when the distance between the data for repair and the data at the current moment is too large and exceeds the threshold l, the correlation between the data for repair and the data at the current moment is not high, and the confidence coefficient of the repair data is reduced.
Preferably, if α is0High or middle or low, the consecutive anomaly count mabnormalSetting 0; otherwise, mabnormal＝mabnormal+1(mabnormalInitially 0). If mabnormalAnd (m is a set parameter) is more than or equal to m, an alarm is given to prompt a user that data is seriously lost, the data is failed to be repaired, and the time point of the data repairing failure and the time interval length mxdeltat of the data repairing failure are output.
Therefore, the invention has the following beneficial effects: (1) the invention provides a method for diagnosing and repairing traffic time sequence data, which can diagnose, mark and repair traffic data from different sources and output marked and more reliable data; (2) after externally input traffic time sequence data are taken, a series of operations such as diagnosis, restoration and marking are carried out on original data, more complete and accurate data are finally output, and the application effect of the data is improved;
drawings
FIG. 1 is a flow chart of the present invention
Detailed Description
The invention is further described with reference to the following detailed description and accompanying drawings.
Example (b): a method for diagnosing and repairing traffic time series data, as shown in fig. 1: the method comprises the following steps:
s1: starting a data preprocessing algorithm to collect data; the method comprises the following specific steps: and after the data preprocessing algorithm is started, starting to import data from each data import component. After externally input traffic time sequence data are taken, a series of operations such as diagnosis, restoration and marking are carried out on original data, more complete and accurate data are finally output, and the application effect of the data is improved;
s2: analyzing and comparing historical data with current data to realize data diagnosis;
s21: acquiring data x corresponding to the current moment; if no data at the current moment exists, marking the abnormal grade p of the data as lose, and directly entering the S3 repairing process;
s22: inquiring and acquiring current-time forward data and current-time historical data of the input data in Step1, wherein the acquisition quantity of the forward data and the historical data is configurable; diagnosing the current time data according to the current time forward data and the current time historical data;
s23: taking a union set of the current time forward data and the current time historical data, calculating an upper quartile Q1, a lower quartile Q3 and a quartile distance IQR (Q3-Q1) of the union set, and if the current time data x is in a range of (Q3-1.5IQR, Q1+1.5IQR)If p is normal; otherwise, entering the calculation of the abnormal degree. If 1.5IQR < Q3-x < alpha2Iqr or 1.5IQR < x-Q1 < alpha1IQR, then p is slot; if α is1IQR＜Q3-x＜α2Iqr or alpha1IQR＜x-Q1＜α2IQR, then p ═ middle; otherwise p is severe. Wherein alpha is1，α2Are all configurable parameters;
s3: using historical data to carry out data restoration on current data;
s31: repairing data with abnormal grade p ═ segment or p ═ lose; the process of repairing the data with the abnormal level p (segment) or p (lose) in step S31 is as follows: inquiring and acquiring data with the abnormal degree lower than midle in the current time forward data and the current time historical data of the input data in the S1; if certain calendar history data is missing, the following steps are executed:
a) data before and after the history data at the time Δ t exist: the average value of Δ t moments before and after the history data is used for replacing:
xnfor missing historical data, xn-1History data at a time Δ t immediately before the missing history data, xn+1History data at a delta t moment after the history data is lost; recording the distance L between the substitute value and the current time datarIs the original distance LoSquare of sum of squares with Δ t/2:
b) data loss at Δ t before and after the history data: replacing data at a delta t moment before the current moment; recording the distance L between the substitute value and the current time datarIs the original distance LoAnd the square sum of Δ t to the square:
the method can process traffic time sequence data of various different sources and types, diagnose the data, and repair the data which is judged to be missing or has higher abnormal grade after diagnosisAnalyzing the confidence level of the repaired data, and finally outputting the more reliable data with the mark
S32: sequencing the acquired forward data at the current time and the historical data at the current time from small to large, taking the first j data (j is a set parameter) with the smallest distance, taking the weighted average of the data as repair data, and taking the weight as the reciprocal of the distance between the data and the data at the current time:
wherein xiFor obtaining current time forward data or current time historical data, LiIs corresponding to xiAnd the distance between the current time data;
s33: setting confidence coefficient, and prompting user data by using the confidence coefficient; the procedure of setting the confidence level in step S33 is: noting the degree of confidence alpha0High. Calculating the number k of the acquired forward data at the current time and the current time historical data, and calculating the maximum distance L between the acquired data and the current time data; if b is not more than k and is less than c, then alpha0If L > L, then alpha is determined0Or else α0Midle; if a is not more than k and is less than b, then alpha0If L > L, then alpha is determined0False, otherwise α0Low; if 0 < k < a, then alpha0False; if k is 0, then α0Fail; otherwise alpha0High; wherein a, b, c and l are all set parameters. c is a first threshold of the repair number, and when the number of the data used for repair is larger than or equal to c, the data used for repair is considered to be enough in quantity, and the repair is reliable; when the number of the data used for repairing is less than c, the data used for repairing is considered to be insufficient, and the confidence coefficient of the repaired data is reduced; b is a second threshold of the repair number, and when the number of the data used for repair is less than b, the data used for repair is considered to be less in quantity, and the confidence of the repair data is continuously reduced; a is a third threshold of the number of repairs, and when the number of repairs is less than a, the repairs can be considered to beThe data volume is too small, and the repair is unreliable; and l is a repair length threshold, and when the distance between the data for repair and the data at the current moment is too large and exceeds the threshold l, the correlation between the data for repair and the data at the current moment is not high, and the confidence coefficient of the repair data is reduced. .
If α is0High or middle or low, the consecutive anomaly count mabnormalSetting 0; otherwise, mabnormal＝mabnormal+1(mabnormalInitially 0). If mabnormalAnd (m is a set parameter) is more than or equal to m, an alarm is given to prompt a user that data is seriously lost, the data is failed to be repaired, and the time point of the data repairing failure and the time interval length mxdeltat of the data repairing failure are output.
Interpretation of terms:
traffic time series data: the data sequence recorded by a certain traffic flow index according to the time sequence identifies the acquisition time of the data through a time stamp field in the data. Intersection lane flow data, for example, is a typical traffic sequence data.
Δ t: two calculated time intervals (default to 5 minutes, configurable).
Forward data at the current time: refers to data several Δ t times before the current time. For example, if the current time is 08:00 and Δ t is configured to be 5 minutes, the previous 07:55, 07:50, and 07:45 data taken forward all the time are forward data.
Current time history data: refers to the data of the current time of the previous weeks. For example, if the current time is wednesday 08:00, the data taken all the way ahead on wednesday 08:00 and wednesday 08:00 are historical data.
Data anomaly level p: a parameter marking the level of data anomalies. p is initially min _ value. If the data is missing, p is normal, if the data is slightly abnormal, p is sleep, if the data is moderately abnormal, p is middle, and if the data is highly abnormal, p is severe.
Distance calculation formula of different date and time data: setting the time interval of two data as x minutes and the date interval as y days, the distance between the two data
Confidence degree alpha0: the marking data may be a parameter of the degree of confidence. Alpha is the higher the confidenceoConfidence is generally αoMiddle, lower confidence is αoα is unknownoFailure of repair is αo＝fail。
Anomaly persistence parameter kabnormal: the duration of the data repair exception is recorded, with larger values for longer durations of repair exceptions.
Claims (9)
1. A method for diagnosing and repairing traffic time series data is characterized by comprising the following steps:
s1: starting a data preprocessing algorithm to collect data;
s2: analyzing and comparing historical data with current data to realize data diagnosis;
s3: and performing data repair on the current data by using the historical data.
2. The method for diagnosing and repairing traffic sequence data according to claim 1, wherein the step S1 comprises the following steps: and after the data preprocessing algorithm is started, starting to import data from each data import component.
3. The method for diagnosing and repairing traffic sequence data according to claim 1, wherein the step S2 includes the steps of:
s21: acquiring data x corresponding to the current moment;
s22: and querying and acquiring current-time forward data and current-time historical data of the input data in the step S1, wherein the acquisition quantity of the forward data and the historical data is configurable.
4. Diagnosing the current time data according to the current time forward data and the current time historical data;
s23: for the forward data at the current timeAnd a current time historical data collection, calculating an upper quartile Q1, a lower quartile Q3 and a quartile distance IQR (Q3-Q1) of the collection, and if the current time data x is in a range of (Q3-1.5IQR, Q1+1.5IQR), determining that the data abnormity grade p is normal; otherwise, entering the calculation of abnormal degree; if 1.5IQR < Q3-x < alpha2Iqr or 1.5IQR < x-Q1 < alpha1IQR, then p is slot; if α is1IQR＜Q3-x＜α2Iqr or alpha1IQR＜x-Q1＜α2IQR, then p ═ middle; otherwise, p is severe; wherein alpha is1，α2Are all configurable parameters.
5. The method as claimed in claim 3, wherein in step S21, if there is no current time data, the flag p ═ lose is marked, and the process proceeds directly to S3 repair process.
6. The method for diagnosing and repairing traffic sequence data according to claim 1, wherein the step S3 includes the steps of:
s31: repairing data with abnormal grade p ═ segment or p ═ lose; s32: sequencing the acquired forward data at the current time and the historical data at the current time from small to large, taking the first j data (j is a set parameter) with the smallest distance, taking the weighted average of the data as repair data, and taking the weight as the reciprocal of the distance between the data and the data at the current time:
wherein xiFor obtaining current time forward data or current time historical data, LiIs corresponding to xiAnd the distance between the current time data;
s33: and setting confidence coefficient, and prompting the user data by using the confidence coefficient.
7. The method for diagnosing and repairing traffic sequence data according to claim 5, wherein the step S31 of repairing the data with abnormal level p (segment) or p (lose) comprises: inquiring and acquiring data with the abnormal degree lower than midle in the current time forward data and the current time historical data of the input data in the S1; if certain calendar history data is missing, the following steps are executed:
a) data before and after the history data at the time Δ t exist: the average value of Δ t moments before and after the history data is used for replacing:
xnfor missing historical data, xn-1History data at a time Δ t immediately before the missing history data, xn+1History data at a delta t moment after the history data is lost; recording the distance L between the substitute value and the current time datarIs the original distance LoSquare of sum of squares with Δ t/2:
b) data loss at Δ t before and after the history data: taking data at a delta t moment before the current moment (the line spacing and other format data can be brushed by a uniform format template when the national bureau of knowledge is submitted), and replacing the data; recording the distance L between the substitute value and the current time datarIs the original distance LoAnd the square sum of Δ t to the square:
8. the method for diagnosing and repairing traffic sequence data according to claim 5, wherein the step of setting the confidence level in step S33 comprises: noting the degree of confidence alpha0High; calculating the number k of the acquired forward data at the current time and the current time historical data, and calculating the maximum distance L between the acquired data and the current time data; if b is not more than k and is less than c, then alpha0If L > L, then alpha is determined0Or else α0Midle; if a is not more than k and is less than b, then alpha0If L > L, then alpha is determined0False, otherwise α0Low; if 0 < k < a, then alpha0False; if k is 0, then α0Fail; otherwise alpha0High; wherein a, b, c and l are all set parameters; c is a first threshold of the repair number, and when the number of the data used for repair is larger than or equal to c, the data used for repair is considered to be enough in quantity, and the repair is reliable; when the number of the data used for repairing is less than c, the data used for repairing is considered to be insufficient, and the confidence coefficient of the repaired data is reduced; b is a second threshold of the repair number, and when the number of the data used for repair is less than b, the data used for repair is considered to be less in quantity, and the confidence of the repair data is continuously reduced; a is a third threshold of the repair number, and when the number used for repair is smaller than a, the repair is unreliable due to the fact that the data volume used for repair is too small; and l is a repair length threshold, and when the distance between the data for repair and the data at the current moment is too large and exceeds the threshold l, the correlation between the data for repair and the data at the current moment is not high, and the confidence coefficient of the repair data is reduced.
9. The method of claim 7, wherein if α is0High or middle or low, the consecutive anomaly count mabnormalSetting 0; otherwise, mabnormal＝mabnormal+1(mabnormalInitially 0);
if mabnormalAnd (m is a set parameter) is more than or equal to m, an alarm is given to prompt a user that data is seriously lost, the data is failed to be repaired, and the time point of the data repairing failure and the time interval length mxdeltat of the data repairing failure are output.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011338594.9A CN112380206B (en) | 2020-11-25 | 2020-11-25 | Diagnosis and repair method of traffic time sequence data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011338594.9A CN112380206B (en) | 2020-11-25 | 2020-11-25 | Diagnosis and repair method of traffic time sequence data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112380206A true CN112380206A (en) | 2021-02-19 |
| CN112380206B CN112380206B (en) | 2022-07-19 |
Family
ID=74588422
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011338594.9A Active CN112380206B (en) | 2020-11-25 | 2020-11-25 | Diagnosis and repair method of traffic time sequence data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112380206B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113190997A (en) * | 2021-04-29 | 2021-07-30 | 贵州数据宝网络科技有限公司 | Big data terminal data restoration method and system |
| CN114944057A (en) * | 2022-04-21 | 2022-08-26 | 中山大学 | Road network traffic flow data restoration method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156579A (en) * | 2014-07-31 | 2014-11-19 | 江南大学 | Dynamic traffic abnormal data detection and recovery method |
| CN110320892A (en) * | 2019-07-15 | 2019-10-11 | 重庆邮电大学 | The sewage disposal device fault diagnosis system and method returned based on Lasso |
| CN111179591A (en) * | 2019-12-30 | 2020-05-19 | 银江股份有限公司 | Road network traffic time sequence characteristic data quality diagnosis and restoration method |
| CN111461564A (en) * | 2020-04-08 | 2020-07-28 | 湖南大学 | Wind turbine generator power characteristic evaluation method based on cloud model and optimal combined weighting |
-
2020
- 2020-11-25 CN CN202011338594.9A patent/CN112380206B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156579A (en) * | 2014-07-31 | 2014-11-19 | 江南大学 | Dynamic traffic abnormal data detection and recovery method |
| CN110320892A (en) * | 2019-07-15 | 2019-10-11 | 重庆邮电大学 | The sewage disposal device fault diagnosis system and method returned based on Lasso |
| CN111179591A (en) * | 2019-12-30 | 2020-05-19 | 银江股份有限公司 | Road network traffic time sequence characteristic data quality diagnosis and restoration method |
| CN111461564A (en) * | 2020-04-08 | 2020-07-28 | 湖南大学 | Wind turbine generator power characteristic evaluation method based on cloud model and optimal combined weighting |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113190997A (en) * | 2021-04-29 | 2021-07-30 | 贵州数据宝网络科技有限公司 | Big data terminal data restoration method and system |
| CN114944057A (en) * | 2022-04-21 | 2022-08-26 | 中山大学 | Road network traffic flow data restoration method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112380206B (en) | 2022-07-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103488881B (en) | Equipment residual service life prediction method under the condition of uncertain degradation measured data | |
| CN112380206B (en) | Diagnosis and repair method of traffic time sequence data | |
| CN110209150B (en) | Job shop scheduling scheme robustness measuring method based on multi-process fault influence | |
| CN111191726B (en) | Fault classification method based on weak supervision learning multilayer perceptron | |
| CN109917777B (en) | Fault detection method based on mixed multi-sampling rate probability principal component analysis model | |
| CN115277464A (en) | Cloud network change flow anomaly detection method based on multi-dimensional time series analysis | |
| KR101908865B1 (en) | Method for data quality analysis of observed temperature | |
| CN111324105B (en) | Small sensor fault detection and separation method based on statistic Mahalanobis distance | |
| CN103618651A (en) | Network abnormality detection method and system based on information entropy and sliding window | |
| CN111898068A (en) | Anomaly detection method based on PERT algorithm and instrument usage analysis | |
| WO2020166236A1 (en) | Work efficiency evaluating method, work efficiency evaluating device, and program | |
| CN115081647A (en) | Industrial intelligent instrument fault pre-diagnosis method based on Bayesian network model | |
| CN112016208B (en) | Hidden fault diagnosis method and system considering disturbance | |
| CN117708735A (en) | Measurement data fitting method, medium and system based on multipoint checking | |
| CN112345972B (en) | Power distribution network line transformation relation abnormity diagnosis method, device and system based on power failure event | |
| CN113010394A (en) | Machine room fault detection method for data center | |
| CN110455370B (en) | Flood-prevention drought-resisting remote measuring display system | |
| CN117195451A (en) | Bridge monitoring data restoration method based on graph theory | |
| CN116933136A (en) | Online ecological observation data anomaly detection method and system | |
| CN110956330A (en) | Method and system for predicting line loss of power transmission line based on multidimensional influence quantity | |
| CN112651444B (en) | Self-learning-based non-stationary process anomaly detection method | |
| CN115809805A (en) | Power grid multi-source data processing method based on edge calculation | |
| Delmas et al. | Remaining useful life estimation methods for predictive maintenance models: defining intervals and strategies for incomplete data | |
| CN113325824B (en) | Regulating valve abnormity identification method and system based on threshold monitoring | |
| CN106547690A (en) | Based on the Software Reliability Modeling method and device under new fitting criterion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 310053 23-25, 2 building, 352 BINKANG Road, Binjiang District, Hangzhou, Zhejiang. Applicant after: Zhejiang zhongkong Information Industry Co.,Ltd. Address before: 310053 23-25, 2 building, 352 BINKANG Road, Binjiang District, Hangzhou, Zhejiang. Applicant before: ZHEJIANG SUPCON INFORMATION TECHNOLOGY Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |