CN112015778A - Water fingerprint prediction algorithm - Google Patents
Water fingerprint prediction algorithm Download PDFInfo
- Publication number
- CN112015778A CN112015778A CN202010838601.5A CN202010838601A CN112015778A CN 112015778 A CN112015778 A CN 112015778A CN 202010838601 A CN202010838601 A CN 202010838601A CN 112015778 A CN112015778 A CN 112015778A
- Authority
- CN
- China
- Prior art keywords
- data
- water
- prediction
- algorithm
- future
- 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.)
- Pending
Links
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 80
- 238000000034 method Methods 0.000 claims abstract description 48
- 238000001914 filtration Methods 0.000 claims abstract description 15
- 238000005192 partition Methods 0.000 claims abstract description 9
- 238000012417 linear regression Methods 0.000 claims abstract description 7
- 230000005540 biological transmission Effects 0.000 claims abstract description 6
- 238000004364 calculation method Methods 0.000 claims description 16
- 230000002194 synthesizing effect Effects 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 6
- 230000002159 abnormal effect Effects 0.000 abstract description 5
- 230000008859 change Effects 0.000 abstract description 3
- 238000012544 monitoring process Methods 0.000 abstract description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- 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
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2474—Sequence data queries, e.g. querying versioned data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
The invention relates to the technical field of water supply network management, and discloses a water fingerprint prediction algorithm, which comprises the following steps: 1) acquiring historical data of a partition needing to be calculated; 2) predicting data for 24 hours in the future by using a time series method; 3) calculating data of the future 24 hours by using a moving average mode; 4) calculating data of the future 24 hours by using a linear regression mode; 5) integrating actual collected data and the two calculated data by using a Kalman filtering method; 6) and calculating different confidence intervals of each algorithm by using a Kalman filtering algorithm. The algorithm calculates the credibility of various algorithms through Kalman filtering, carries out comprehensive prediction, can cope with various different special conditions, has wider application range, selects concentrated basic algorithms with different characteristics, adapts to different scenes, avoids abnormal change of returned data possibly caused by various reasons and influences on normal monitoring and analysis of the data in water meter flow data acquired by remote transmission equipment.
Description
Technical Field
The invention relates to the technical field of water supply network management, in particular to a water fingerprint prediction algorithm.
Background
The water fingerprint prediction algorithm belongs to water supply network management and can be applied to the field of subarea measurement and leakage control.
At present, for the data collected by a flowmeter or a water meter in the partition metering, due to reasons such as equipment failure or signal abnormality, the accuracy of the data can not be guaranteed by 100%, the possibility of data missing and burrs exists, the interference of the bad data is easily received, the normal analysis of the monitored data is influenced, for example, false alarm and the like are generated due to burrs, although the prior art can process the bad data, the processing method is simpler, the adaptability is not wide enough, and the following problems mainly exist: 1. directly removing the burr data, wherein the data at the moment is empty after the burr data are removed, and the continuity of the data is influenced although no burr exists; 2. the compensation for the missing data is not intelligent enough, for example, the average value is simply calculated, and the compensation effect is poor compared with the real data; 3. because the conditions of data loss, abnormality and the like influence the implementation of algorithms such as subsequent prediction, alarm and the like, the water fingerprint prediction algorithm is provided for solving the problems
Disclosure of Invention
Technical problem to be solved
Aiming at the defects of the prior art, the invention provides a water fingerprint prediction algorithm, which has the advantages of judging a prediction interval with reasonable water quantity of a subarea or a meter, assisting workers in judging the running state of the subarea or the meter, realizing timely pre-judgment and timely alarm, being used as a basis for leakage control work and the like, and solves the problems of influencing the continuity of data, greatly distinguishing from real data, influencing the follow-up prediction and alarm and other algorithms.
(II) technical scheme
In order to achieve the purpose, the invention provides the following technical scheme: a prediction interval for judging the reasonable water quantity of a partition or a meter is formed mainly by comprehensively analyzing historical data and prediction data and adopting a machine learning method, the operation state of the partition or the meter is judged by auxiliary workers, the timely prejudgment and the timely alarm are realized, and the prediction interval can be used as the basis of leakage control work, and the specific method comprises the following steps:
a water fingerprint prediction algorithm comprising the steps of:
1) acquiring historical data of a partition needing to be calculated, and directly calculating water quantity data through values acquired by remote transmission equipment;
2) the time series method is used for predicting data of 24 hours in the future, and most of water using data have certain periodicity, so that the time series method can be used for summarizing and analyzing historical data and predicting the water amount of a period of time in the future;
3) calculating data of 24 hours in the future by using a moving average mode, wherein most of user water quantities have daily relevance, and water quantities at the same time on different days have obvious relevance, and the method estimates the water quantity data of 24 hours in the future by averaging the data of the water at the moment of nearly 7 days;
4) the method comprises the steps that data of the water meter in the future 24 hours are calculated in a linear regression mode, the water meter data can be used as a prediction basis through other related data besides the water meter data, the method uses the factory flow of a plurality of water plants as a calculation basis, the historical data are fitted in a linear regression method, and the water quantity of the water meter at the next moment is predicted;
5) the method comprises the steps that actually acquired data and the two calculation data are integrated by using a Kalman filtering method, final water fingerprint data are obtained through integrated calculation, various prediction methods have certain errors, the Kalman filtering method can be adopted, the variance of normal distribution of prediction errors in a period of time of each prediction method is calculated and used as the credibility of the prediction method, and more credible prediction results can be obtained through integrating the prediction results of the 3 different calculation methods;
6) and calculating different confidence intervals of each algorithm by using a Kalman filtering algorithm, and synthesizing each algorithm to obtain a final prediction result, wherein the final result has the highest adaptability.
Preferably, the water amount data calculation formula in step 1) is as follows: the intermittent water amount is the current time accumulated value-the last time accumulated value.
Preferably, the time series method in step 2) is implemented according to facebook open source fbprophet.
Preferably, the variance calculation formula of each method error in the step 5) is as follows:
where N represents the number of samples, μ represents the average of the sample set, and Xn represents the nth sample.
(III) advantageous effects
Compared with the prior art, the invention provides a water fingerprint prediction algorithm, which has the following beneficial effects:
the water fingerprint prediction algorithm is characterized in that historical data of a terminal are studied by comprehensively utilizing various algorithms to generate simulated prediction data, so that the influence caused by data loss is made up, meanwhile, a normal reference curve interval of each water meter can be provided, the operation condition of the terminal is assisted to be judged manually, abnormal conditions are found in time, the algorithm is different from the limitation of a common single prediction algorithm, the reliability of various algorithms is calculated by Kalman filtering, comprehensive prediction is carried out, various different special conditions can be dealt with, the application range is wider, the selected concentrated basic algorithm has different characteristics and is suitable for different scenes, and the problems that the normal monitoring and analysis of the data are influenced due to abnormal change of returned data caused by various reasons due to the fact that the water meter flow data collected by remote transmission equipment are avoided.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
A water fingerprint prediction algorithm is mainly characterized in that a prediction interval for judging the reasonable water quantity of a partition or a meter is formed by comprehensively analyzing historical data and prediction data and adopting a machine learning method, and workers are assisted in judging the running state of the partition or the meter, so that timely prejudgment and timely alarm are realized, and the prediction algorithm can be used as a basis for leakage control work and comprises the following steps:
1) acquiring historical data of a partition needing to be calculated, and directly calculating water quantity data through values acquired by remote transmission equipment, wherein the water quantity data calculation formula is as follows: the intermittent water quantity is the current time accumulated value-the last time accumulated value;
2) the data of 24 hours in the future are predicted by using a time series method, most of water consumption data have certain periodicity, so that historical data can be summarized and analyzed by using the time series method, the water quantity of a period of time in the future is predicted, and the time series method is realized according to facebook open-source fbprophetet;
3) calculating data of 24 hours in the future by using a moving average mode, wherein most of user water quantities have daily relevance, and water quantities at the same time on different days have obvious relevance, and the method estimates the water quantity data of 24 hours in the future by averaging the data of the water at the moment of nearly 7 days;
4) the method comprises the steps that data of the water meter in the future 24 hours are calculated in a linear regression mode, the water meter data can be used as a prediction basis through other related data besides the water meter data, the method uses the factory flow of a plurality of water plants as a calculation basis, the historical data are fitted in a linear regression method, and the water quantity of the water meter at the next moment is predicted;
5) the method comprises the steps of using a Kalman filtering method to synthesize actual collected data and the two calculation data, obtaining final water fingerprint data through comprehensive calculation, enabling various prediction methods to have certain errors, adopting the Kalman filtering method to calculate the variance of normal distribution of prediction errors in a period of time of each prediction method, using the variance as the credibility of the prediction method, and obtaining more credible prediction results by synthesizing the prediction results of the 3 different calculation methods, wherein the variance calculation formula of errors of each method is as follows:
wherein N represents the number of samples, mu represents the average value of the sample set, and Xn represents the nth sample;
6) and calculating different confidence intervals of each algorithm by using a Kalman filtering algorithm, and synthesizing each algorithm to obtain a final prediction result, wherein the final result has the highest adaptability.
When the method is used, 1, the moving average algorithm is based on data at the same moment in historical data, the precision is very high under the condition that the periodicity of water for a user is strong, but the prediction period is not long;
2. the time sequence method makes up the defect that the moving average algorithm is not processed on special dates and can predict longer time based on the relation between historical data and time, including the influence of weekends, holidays and the like on water use;
3. the moving average algorithm carries out prediction analysis according to the linear relation between the actual water yield of the water source and the equipment, is equivalent to a water-using macroscopic model, improves the prediction precision through certain basic data, but has shorter prediction time;
4. aiming at different characteristics of the algorithms, respective credibility is calculated through Kalman filtering to generate a comprehensive prediction result, so that the method is suitable for more application scenes.
The invention has the beneficial effects that: the water fingerprint prediction algorithm is characterized in that historical data of a terminal are learned by comprehensively utilizing various algorithms to generate simulated prediction data, so that the influence caused by data loss is compensated, a normal reference curve interval of each water meter can be provided, the operation condition of the terminal is assisted to be judged manually, abnormal conditions are found timely, the method is different from the limitation of a common single prediction algorithm, the reliability of various algorithms is calculated by Kalman filtering, comprehensive prediction is carried out, various different special conditions can be dealt with, the application range is wider, the selected concentrated basic algorithms have different characteristics and are suitable for different scenes, the problems that the normal monitoring and analysis of data are influenced due to abnormal change of returned data caused by various reasons due to the fact that the water meter flow data collected by remote transmission equipment are avoided, and the problems that the continuity, the operation condition and the analysis of the data are influenced, Compared with the real data, the method has larger difference and influences the subsequent prediction and alarm algorithm.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (4)
1. A water fingerprint prediction algorithm, comprising the steps of:
1) acquiring historical data of a partition needing to be calculated, and directly calculating water quantity data through values acquired by remote transmission equipment;
2) the time series method is used for predicting data of 24 hours in the future, and most of water using data have certain periodicity, so that the time series method can be used for summarizing and analyzing historical data and predicting the water amount of a period of time in the future;
3) calculating data of 24 hours in the future by using a moving average mode, wherein most of user water quantities have daily relevance, and water quantities at the same time on different days have obvious relevance, and the method estimates the water quantity data of 24 hours in the future by averaging the data of the water at the moment of nearly 7 days;
4) the method comprises the steps that data of the water meter in the future 24 hours are calculated in a linear regression mode, the water meter data can be used as a prediction basis through other related data besides the water meter data, the method uses the factory flow of a plurality of water plants as a calculation basis, the historical data are fitted in a linear regression mode, and the water quantity data of the water meter at the next moment are predicted;
5) the method comprises the steps that actually acquired data and the two calculation data are integrated by using a Kalman filtering method, final water fingerprint data are obtained through integrated calculation, various prediction methods have certain errors, the Kalman filtering method can be adopted, the variance of normal distribution of prediction errors in a period of time of each prediction method is calculated and used as the credibility of the prediction method, and more credible prediction results can be obtained through integrating the prediction results of the 3 different calculation methods;
6) and calculating different confidence intervals of each algorithm by using a Kalman filtering algorithm, and synthesizing each algorithm to obtain a final prediction result, wherein the final result has the highest adaptability.
2. The water fingerprint prediction algorithm of claim 1, wherein the water amount data in step 1) is calculated by the following formula: the intermittent water amount is the current time accumulated value-the last time accumulated value.
3. The water fingerprint prediction algorithm of claim 1, wherein the time series method in step 2) is implemented according to facebook open source fbprophet.
4. The water fingerprint prediction algorithm of claim 1, wherein the variance of each method error in step 5) is calculated by the following formula:
where N represents the number of samples, μ represents the average of the sample set, and Xn represents the nth sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010838601.5A CN112015778A (en) | 2020-08-19 | 2020-08-19 | Water fingerprint prediction algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010838601.5A CN112015778A (en) | 2020-08-19 | 2020-08-19 | Water fingerprint prediction algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112015778A true CN112015778A (en) | 2020-12-01 |
Family
ID=73505135
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010838601.5A Pending CN112015778A (en) | 2020-08-19 | 2020-08-19 | Water fingerprint prediction algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112015778A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114593375A (en) * | 2022-03-30 | 2022-06-07 | 常州通用自来水有限公司 | Secondary water supply community pipeline leakage monitoring and positioning method based on pump house energy consumption |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102157075A (en) * | 2011-03-15 | 2011-08-17 | 上海交通大学 | Method for predicting bus arrivals |
| CN102183802A (en) * | 2011-03-10 | 2011-09-14 | 西安交通大学 | Short-term climate forecast method based on Kalman filtering and evolution modeling |
| CN107292410A (en) * | 2016-03-30 | 2017-10-24 | 中国石油天然气股份有限公司 | Tunnel deformation Forecasting Methodology and device |
| CN108766585A (en) * | 2018-05-31 | 2018-11-06 | 平安科技（深圳）有限公司 | Generation method, device and the computer readable storage medium of influenza prediction model |
| CN110287638A (en) * | 2019-07-04 | 2019-09-27 | 南京邮电大学 | Winged saw life-span prediction method based on kalman-RNN neural network |
| CN111242284A (en) * | 2020-01-10 | 2020-06-05 | 支付宝(杭州)信息技术有限公司 | Prediction method and device |
-
2020
- 2020-08-19 CN CN202010838601.5A patent/CN112015778A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102183802A (en) * | 2011-03-10 | 2011-09-14 | 西安交通大学 | Short-term climate forecast method based on Kalman filtering and evolution modeling |
| CN102157075A (en) * | 2011-03-15 | 2011-08-17 | 上海交通大学 | Method for predicting bus arrivals |
| CN107292410A (en) * | 2016-03-30 | 2017-10-24 | 中国石油天然气股份有限公司 | Tunnel deformation Forecasting Methodology and device |
| CN108766585A (en) * | 2018-05-31 | 2018-11-06 | 平安科技（深圳）有限公司 | Generation method, device and the computer readable storage medium of influenza prediction model |
| CN110287638A (en) * | 2019-07-04 | 2019-09-27 | 南京邮电大学 | Winged saw life-span prediction method based on kalman-RNN neural network |
| CN111242284A (en) * | 2020-01-10 | 2020-06-05 | 支付宝(杭州)信息技术有限公司 | Prediction method and device |
Non-Patent Citations (1)
| Title |
|---|
| 董洁等: "《南四湖流域水资源工序格局与对策研究》", 30 April 2015, 黄河水利出版社 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114593375A (en) * | 2022-03-30 | 2022-06-07 | 常州通用自来水有限公司 | Secondary water supply community pipeline leakage monitoring and positioning method based on pump house energy consumption |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10839115B2 (en) | Cleansing system for a feed composition based on environmental factors | |
| CN111307123B (en) | Real-time abnormity diagnosis and interpolation method of regimen monitoring data | |
| US20170369800A1 (en) | Method of Reducing Corrosion and Corrosion Byproduct Deposition in a Crude Unit | |
| US20170315543A1 (en) | Evaluating petrochemical plant errors to determine equipment changes for optimized operations | |
| CN110008565A (en) | A kind of industrial process unusual service condition prediction technique based on operating parameter association analysis | |
| JP2020170003A (en) | Method for reducing corrosion in oil refineries | |
| CN113032454A (en) | Interactive user power consumption abnormity monitoring and early warning management cloud platform based on cloud computing | |
| CN111861206A (en) | Industrial industry prosperity index obtaining method based on enterprise electric power big data | |
| DK180559B1 (en) | CIP control surveillance system and application of the system | |
| CN112015778A (en) | Water fingerprint prediction algorithm | |
| CN113539382B (en) | Early warning positioning method and system for key technological parameters of dimethyl phosphite | |
| CN116073436A (en) | Capacity optimization control method for photovoltaic new energy power system | |
| CN109813782B (en) | Self-diagnosis PH on-line monitoring system and method | |
| CN117290795A (en) | Online identification method suitable for abnormal values of measurement results of single instrument | |
| CN105718620A (en) | Modeling method for reliability statistic analysis of key component of converter valve cooling system | |
| Alferes et al. | Data quality assurance in monitoring of wastewater quality: Univariate on-line and off-line methods | |
| CN115809805A (en) | Power grid multi-source data processing method based on edge calculation | |
| CN111174824B (en) | Control platform that acid mist discharged | |
| CN114962239A (en) | Equipment fault detection method based on intelligent Internet of things | |
| CN110083804B (en) | Wind power plant SCADA data missing intelligent repairing method based on condition distribution regression | |
| CN114970311A (en) | Method for establishing remote module life prediction model and life prediction method | |
| CN110377007A (en) | Fault Diagnosis for Chemical Process method based on pivot analysis | |
| CN114819743B (en) | Energy consumption diagnosis and analysis method for chemical enterprises | |
| CN117034166A (en) | Robust dam monitoring abnormal data online identification method capable of adapting to different departure factors | |
| WO2019005541A1 (en) | Evaluating petrochemical plant errors to determine equipment changes for optimized operations |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201201 |
|
| RJ01 | Rejection of invention patent application after publication |