CN113792988A - Online monitoring data anomaly identification method for enterprise - Google Patents
Online monitoring data anomaly identification method for enterprise Download PDFInfo
- Publication number
- CN113792988A CN113792988A CN202110974638.5A CN202110974638A CN113792988A CN 113792988 A CN113792988 A CN 113792988A CN 202110974638 A CN202110974638 A CN 202110974638A CN 113792988 A CN113792988 A CN 113792988A
- Authority
- CN
- China
- Prior art keywords
- enterprise
- production
- data
- abnormal
- online
- 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
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000012544 monitoring process Methods 0.000 title claims abstract description 46
- 238000004519 manufacturing process Methods 0.000 claims abstract description 172
- 230000002159 abnormal effect Effects 0.000 claims abstract description 82
- 230000005611 electricity Effects 0.000 claims abstract description 47
- 230000000694 effects Effects 0.000 claims abstract description 38
- 238000001514 detection method Methods 0.000 claims abstract description 7
- 238000010219 correlation analysis Methods 0.000 claims abstract description 6
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 239000003344 environmental pollutant Substances 0.000 claims description 50
- 231100000719 pollutant Toxicity 0.000 claims description 50
- 238000012545 processing Methods 0.000 claims description 26
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 20
- 239000001301 oxygen Substances 0.000 claims description 20
- 229910052760 oxygen Inorganic materials 0.000 claims description 20
- 239000002912 waste gas Substances 0.000 claims description 10
- 230000005856 abnormality Effects 0.000 claims description 9
- 239000007789 gas Substances 0.000 claims description 6
- 230000000630 rising effect Effects 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000005452 bending Methods 0.000 claims description 3
- 239000000356 contaminant Substances 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 230000001174 ascending effect Effects 0.000 claims description 2
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 230000006399 behavior Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000003912 environmental pollution Methods 0.000 description 4
- 230000007613 environmental effect Effects 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 239000003546 flue gas Substances 0.000 description 2
- 239000000779 smoke Substances 0.000 description 2
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Theoretical Computer Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
An enterprise online monitoring data anomaly identification method comprises the following steps: 1) acquiring enterprise online data and sub-meter electricity metering data, and preprocessing the data to obtain enterprise activity data; 2) generating a self-adaptive sliding window, and carrying out high-value detection on the enterprise activity data in the window to obtain a high-value sequence and a low-value sequence; 3) performing correlation analysis on the high and low value sequence data obtained in the step 2 to obtain correlation results among the data; 4) calling corresponding judgment logic according to the corresponding enterprise abnormal characteristics; 5) and judging the abnormal conditions and types of the enterprise activities. The power consumption data is introduced, so that the monitoring data are mutually verified, and the reliability of the identification result is improved; a self-adaptive sliding window is generated, so that the identification capability of an abnormal high value is improved, and the result accuracy is improved; establishing judgment logic for online data anomaly identification; the method realizes the identification of the implementation situation of the enterprise production reduction and limitation measures, the real discharge variation situation and the abnormal operation of the production and treatment facilities for the first time.
Description
Technical Field
The invention relates to the field of environmental pollution early warning, in particular to an enterprise online monitoring data abnormity identification method.
Background
In recent years, with the acceleration of economic construction and social development speed in China, environmental problems are increasingly prominent. The development of industrial economy is the main driving force for the high-speed development of economy in China, and the problem of environmental pollution caused by the development of industrial economy is a main component of the environmental problem. The monitoring of exhaust gases from industrial enterprises mainly comprises monitoring of flue gases, i.e. exhaust gases from industrial boilers and kilns and process exhaust gases. Based on the online data and the power consumption data of the industrial enterprise, the pollutant emission characteristics of various industries, production and treatment processes are analyzed, the online data abnormal characteristics of the industrial enterprise are fully excavated, the production emission rules of the industrial enterprise are summarized, the environment monitoring behavior can be effectively standardized, and the pollution problem can be timely found.
However, the process of finding the problem of environmental pollution based on massive monitoring data at present requires a large amount of manpower and time, has high dependence on the technical level and experience of research personnel, has low efficiency in the whole application process, is poor in timeliness and is limited by the level of the technical personnel, and is difficult to effectively support environmental management. The method for identifying the monitoring data abnormity of the industrial enterprise efficiently and quickly is needed for providing support for scientifically monitoring the atmospheric pollution emission of the industrial enterprise and effectively improving the regional air quality.
Many data processing methods have been developed in this industry, such as CN110245880A entitled: the invention relates to a method for identifying cheating of online monitoring data of a pollution source, and the invention name of CN110990393A is as follows: the patent of 'a big data identification method for abnormal data behaviors of industry and enterprise', and the like, however, most of the patents only carry out statistics and analysis on online monitoring data and only identify the counterfeiting behaviors of the monitoring data, the problem of environmental pollution is not involved, and the identification function has certain limitation.
Disclosure of Invention
In order to solve the above problems, the present invention provides a method for identifying abnormality of online monitoring data of an enterprise.
In order to solve the above problems, the present invention provides an enterprise online monitoring data anomaly identification method, which comprises the following steps:
1) acquiring enterprise pollutant parameter data and power consumption parameter data, and preprocessing to obtain enterprise activity data: accessing enterprise pollutant parameter data and power consumption parameter data, and storing by using a datatable; the steps may include:
(1) longitudinally processing pollutant parameter data and electricity consumption parameter data to obtain the exhaust emission (m)3) Smoke concentration (mg/m)3)、SO2Concentration (mg/m)3) NOx concentration (mg/m)3) The system comprises pollutant parameters, electricity consumption parameters such as electricity consumption of production facilities and electricity consumption of treatment facilities, and basic information parameters such as enterprise names, discharge names, production facility names, waste gas treatment facility names, monitoring time, monitoring point names and the like;
(2) matching the treated pollutant parameter data with the electricity utilization parameter data, and screening out the gas-related enterprises to be monitored and corresponding information: the name of the discharge port, the power consumption of a production facility and the power consumption of a processing facility;
(3) transversely processing the online data of the enterprise, screening out enterprise row openings with serious enterprise pollutant parameter data loss and corresponding discharge amount, and removing enterprise row openings with insufficient data in continuous hours, so that enough data can enter the sliding window in the first stage;
(4) and (3) carrying out cubic spline interpolation on the cleaned enterprises and the data of the corresponding exhaust pollutant parameters: a process of forming a smooth curve passing through a series of shape value points and obtaining a curve function set by solving a three bending moment equation set;
(5) and forming pollutant parameter time series activity data corresponding to different discharge ports of different enterprises and production facility power consumption and processing facility power consumption time series activity data matched with the discharge ports.
2) Generating a self-adaptive sliding window, and carrying out high-value detection on the enterprise activity data in the window to obtain a high-value sequence and a low-value sequence;
wherein the adaptive sliding window method identifies high values: the method is characterized in that the window size and the threshold variable setting are adjusted according to different data types of different enterprises and by combining the generated process characteristics and the data change rules expressed in daily production activities, the effective identification characteristic high value is determined, and characteristic sliding windows are established according to different enterprise types;
3) performing correlation analysis on the high and low value sequence data obtained in the step 2 to obtain correlation results among the data;
4) calling corresponding judgment logic according to the corresponding enterprise abnormal characteristics;
in this step, it is preferable that: through analyzing pollutant parameter information such as pollutant concentration, waste gas emission, oxygen content and the like of an enterprise online monitoring system, sub-metering the change conditions of power consumption parameters such as power consumption of production facilities and power consumption of processing facilities of an electric system, combing the logical relationship among the parameters and corresponding abnormal characteristics of an enterprise in the production and production stop processes of the enterprise, and obtaining a judgment logical library of the abnormal characteristics of the enterprise. Calling corresponding judgment logic according to the abnormal characteristics of the enterprise;
according to the production and production stop states marked by the online monitoring system, enterprise activity data are divided into two types, namely production logic and production stop logic, the production logic is called to judge in the production state, and the production stop logic is called to judge in the production stop state;
5) and judging the abnormal conditions and abnormal types of the enterprise activities.
Preferably, the step of determining the abnormal conditions and types of the enterprise activities is as follows:
based on the obtained activity data such as the enterprise pollutant parameters, the electricity utilization parameters and the like, judging logic of enterprise abnormal characteristics is called to analyze whether the enterprise activity is abnormal or not, and the abnormal type is judged (abnormal rising A, abnormal rising B, abnormal production A, abnormal production B and abnormal production stop).
(1) Abnormal rise a (data counterfeiting): the concentration of pollutants does not accord with the logics of the exhaust emission, the power consumption of production facilities, the power consumption of treatment facilities and the like.
(2) Abnormal rise B: the processing measures do not match the production load;
production abnormity: and data belonging to the characteristic production load are not matched, namely the exhaust emission of the enterprise is inconsistent with the power consumption of the production facility.
(3) Production anomaly A: the amount of exhaust emissions is abnormally low.
(4) Production anomaly B: the amount of exhaust emissions is abnormally high.
(5) Production stoppage abnormity:
and if parameters such as oxygen content, pollutant concentration, exhaust emission, electricity consumption of production facilities and the like in the production stopping process have any conditions which do not accord with normal production stopping conditions, judging that the production stopping is abnormal.
The invention has the beneficial effects that the technical scheme has the following advantages:
1. on the basis of the pollutant parameters, electricity utilization parameters are introduced, so that the monitoring data of two relatively independent systems are mutually proved, and the reliability of the identification result is improved;
2. the self-adaptive sliding window is generated, high-value detection is carried out on the enterprise activity data in the window, the identification capability of abnormal high values is improved, and the result accuracy is improved;
3. based on case study, enterprise abnormal emission behaviors are summarized, manual judgment experience is extracted, and judgment logic of online data abnormal recognition is established;
4. the method realizes the identification of the implementation situation of the enterprise production reduction and limitation measures, the real discharge variation situation and the abnormal operation of the production and treatment facilities for the first time.
Drawings
Fig. 1 is a flowchart of an enterprise online monitoring data anomaly identification method according to the present invention.
FIG. 2 is a schematic diagram of a smooth curve formed by Cubic Spline Interpolation on cleaned enterprises and corresponding data with missing nozzle emphasis parameters.
Detailed Description
The technical scheme of the invention is further specifically described by the following embodiments and the accompanying drawings.
The idea of the invention and the related technical characteristic data mainly comprise:
the data source is enterprise activity data which comprises pollutant parameters, state parameters, electricity utilization parameters and basic information parameters. Wherein the pollutant parameter includes, but is not limited to, an amount of exhaust emissions (m)3) Smoke concentration (mg/m)3)、SO2Concentration (mg/m)3) NOx concentration (mg/m)3) TVOC (ppm); status parameters including, but not limited to, temperature, pressure, flow rate, oxygen content, humidity of the flue gas, and production/off-stream status of the production facility, production/off-stream status of the exhaust treatment facility; electricity consumption parameters, electricity consumption of production facilities and electricity consumption of processing facilities; basic information parameters, enterprise names, discharge port names, production facility names, waste gas treatment facility names, monitoring point names and monitoring time.
And (3) an anomaly identification algorithm: acquiring enterprise activity data through online data acquisition and pretreatment based on pollutant parameters and state parameters of enterprise online data and electricity consumption parameters such as electricity consumption of production facilities and electricity consumption of treatment facilities; carrying out high-value detection on the enterprise activity data by using a self-adaptive sliding window algorithm to obtain a high-value sequence and a low-value sequence; performing correlation analysis on the high-low value sequence to obtain correlation results among the data; and calling corresponding judgment logics according to the corresponding abnormal characteristics of the enterprises, and judging whether the enterprises have abnormal emission conditions or not by combining production processes and treatment process characteristics of different industries.
The function is as follows: identifying enterprise data counterfeiting, pushing the enterprise data counterfeiting to an environment management department, and assisting law enforcement and supervision; identifying the implementation situation of production reduction measures, pushing the implementation situation to an environment management department, and assisting law enforcement and supervision; the emission variation is found, and the air quality pollution source tracing is supported; and (4) finding out the facility abnormality, pushing the abnormality to the enterprise, assisting the process optimization and the facility maintenance of the enterprise, and supporting the energy conservation and emission reduction of the enterprise.
Referring to fig. 1, a flowchart of an enterprise online monitoring data anomaly identification method according to the present invention is shown, which mainly includes the following steps:
1) starting;
2) accessing online data and sub-meter electricity metering data of an enterprise, and preprocessing the online data and the sub-meter electricity metering data to obtain enterprise activity data;
3) generating a self-adaptive sliding window, and carrying out high-value detection on the enterprise activity data in the window to obtain a high-value sequence and a low-value sequence;
4) carrying out correlation analysis on the high and low value sequence data to obtain a correlation result;
5) calling corresponding judgment logic according to the corresponding enterprise abnormal characteristics;
6) judging whether the enterprise activity data belongs to an abnormity or not, and giving out an abnormity category;
7) and (6) ending.
The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
In this embodiment, the method for identifying abnormality of online data of an enterprise of the present invention includes the following specific steps:
1. accessing online data and sub-meter electricity metering data of an enterprise, and preprocessing to obtain enterprise activity data:
accessing online data of an enterprise and sub-meter electricity metering data, and storing by using a datatable:
(1) longitudinally processing online data of enterprises to obtain SO (SO) changing with hours2Monitoring pollutant parameters such as concentration data, NOx monitoring concentration data and exhaust emission, and corresponding basic information parameters such as enterprise name, discharge port name, monitoring time, oxygen content and whether to shut down; and longitudinally processing the sub-meter electricity metering data to obtain electricity utilization parameters such as power of production facilities and processing facilities and basic information parameters such as enterprise names, facility types, monitoring point names and acquisition time.
(2) And matching the processed online data of the enterprises with the sub-meter electricity metering data, and screening the gas-related enterprises to be monitored and corresponding information such as corresponding discharge openings, the electricity consumption of production facilities, the electricity consumption of processing facilities and the like.
(3) Horizontal processing enterprise online data
Screening of contaminant parameters (SO)2Or NOx and exhaust emission) and corresponding emission (the data of pollutant parameters are missing more than 80%), removing the enterprise exhaust with too few data in continuous hours (removing the enterprise exhaust with too few lines, such as the line number less than 100), and ensuring that enough data enters the second-step sliding window in the early stage.
(4) And (3) carrying out Cubic Spline Interpolation on the cleaned enterprises and the data with the missing corresponding discharge pollutant parameters in the steps: obtaining a smooth curve passing through a series of shape value points, and then obtaining a curve function set by solving a three bending moment equation set.
And after the interpolation is completed, pollutant parameter time series activity data corresponding to different discharge ports of different enterprises and electric power time series activity data of production facilities and processing facilities matched with the discharge ports are formed.
2. Generating a self-adaptive sliding window, carrying out high-value detection on the enterprise activity data in the window to obtain a high-value sequence and a low-value sequence,
the adaptive sliding window method identifies high values: the method is characterized in that the window size and the threshold variable setting are adjusted according to different data types of different enterprises and the data change rules expressed in the daily production activities, so that the effective identification characteristic high value is determined. A sliding window of features is established for different business types.
And (3) calculating the activity data in the step (1) in a sliding window: setting an initial sliding window size to be n (setting of the sliding window size needs to match the characteristic condition of the active data vector), setting a signal vector signals to be 0,.. 0 (length is the active data vector length L), setting an initial sliding window size variable n, a threshold variable threshold, a peak clipping variable influ, setting an active data vector filtered y in the sliding window to be y (1), and.
Standard deviation within the current window:
the absolute value of the difference between the first data y (n + i) (i 1.. L-n) outside the current window and the mean of the active data within the current sliding window:
diff(n+i)＝|y(n+i)-avgFilter(n)| (3)
the fluctuation amount of the data vector in the window is as follows:
fluctuate(n)＝threshold*stdFilter(n) (4)
if diff (n + i) > fluctuate and y (n + i) > avgfilter (n) are satisfied, the signal value of the n + i-th data is set to 1 (peak signal), and if diff (n + i) > fluctuate and y (n + i) < avgfilter (n) at this time, the signal value of the n + i-th data is set to-1 (valley signal), the above process is followed by peak-clipping the filteredY (n + i) variable:
filteredY(n+i)＝influence*y(n+i)+(1-influence)*filteredY(n)(5)
if diff (n + i) < yield is satisfied, the signal value of the n + i-th data is set to 0 (wave-level signal), and filteredY (n + i) is assigned to y (n + i):
filteredY(n+i)＝y(n+i) (6)
and (3) cyclic calculation: and moving the whole window backward by one unit data, and calculating values (1) to (6) of a new window to obtain a signal value of the next data until the calculation of the data vector with the length of L is completed and a final signal vector signals is obtained.
3. And (4) carrying out correlation analysis on the high-low value sequence data obtained in the step 2 to obtain a correlation result among the data.
4. Calling corresponding judgment logic according to the corresponding enterprise abnormal characteristics,
through analyzing the change conditions of power consumption parameters such as the actually measured pollutant concentration, the exhaust emission and the oxygen content of an enterprise online monitoring system, the power consumption of production facilities of an electric system, the power consumption of processing facilities and the like, the logical relationship among the parameters and the corresponding abnormal characteristics of the enterprise in the production and production stop processes of the enterprise are combed, and a judgment logical library of the abnormal characteristics of the enterprise is obtained. And calling corresponding judgment logic according to the abnormal characteristics of the enterprise.
According to the production and shutdown states marked by the online monitoring system, enterprise activity data is divided into two types, namely production logic and shutdown logic, the production logic is called to judge in the production state, and the shutdown logic is called to judge in the shutdown state.
(1) Enterprise production logic judgment:
the first step is as follows: and judging whether the change trends of the hour data of the power consumption of the production facilities and the exhaust emission of the enterprise are changed synchronously (ascending/unchanging/reducing), if not, judging that the changes of the two factors of the exhaust emission and the power consumption of the facility representing the production load are inconsistent, and judging that the changes are not in accordance with the production logic of the enterprise. If synchronous, then enter the second step.
The second step is that: and judging whether the change trends of the power consumption of the processing facility and the exhaust emission are changed synchronously or not, if so, judging that the production and emission logic of the enterprise is met, and if not, judging that the production and emission logic is not met.
The third step: and (4) judging the abnormal emission type by analyzing the variation relation among the parameters, and entering the steps (1) - (4) of the step 5.
(2) Enterprise production stopping logic judgment:
the first step is as follows: and when the oxygen content is 19.5-22.0%, the pollutant concentration is far lower than normal production emission (< average of historical production 20%), the exhaust emission is 0 or far lower than normal production emission (< average of historical production 20%), and the production power consumption is 0 or far lower than normal power consumption level (< average of historical production 20%), the normal production stop is judged.
The second step is that: and if any condition which does not accord with the normal production halt appears in the production halt process, judging the abnormal production halt. Note that: the process of shutdown or startup can be carried out 5 hours after the shutdown and 5 hours before the shutdown, and the oxygen content, the exhaust emission and the pollutant concentration can fluctuate (each parameter is higher) in a reducing or increasing way, which belongs to the normal situation.
The third step: and (5) judging the type of the abnormal production halt by analyzing the data change conditions of the oxygen content, the pollutant concentration, the exhaust emission and the electricity consumption of the production facility, and entering the step 5.
5. Judging the abnormal condition and type of the enterprise activities,
based on the obtained activity data such as the online data of the enterprise, the electricity consumption and the like, judging logic of enterprise abnormal characteristics is called to analyze whether the enterprise activity is abnormal or not, and the abnormal type is judged (abnormal rising A, abnormal rising B, abnormal production A, abnormal production B and abnormal production stop).
(1) Abnormal rise a (data counterfeiting): the concentration of pollutants does not accord with the logics of the exhaust emission, the power consumption of production facilities, the power consumption of treatment facilities and the like.
When the data of the waste gas emission amount of the enterprise and the electricity consumption amount of the production facility are unchanged, the electricity consumption amount of the processing facility is increased, the pollutant concentration is increased, and the abnormal increase A is judged.
And secondly, when the data of the exhaust emission of the enterprise and the power consumption of the production facility are reduced, the power consumption of the processing facility is unchanged, and the concentration of the pollutants is increased, judging that the pollutant is abnormally increased A.
(2) Abnormal rise B: the treatment measures do not match the production load.
When the exhaust emission of an enterprise increases, the power consumption of production facilities increases, but the power consumption of treatment facilities is unchanged or decreases, so that the concentration of pollutants increases, and the concentration is judged to be abnormally increased B.
And secondly, when the exhaust emission of the enterprise and the power consumption of the production facility are unchanged, the power consumption of the treatment facility is reduced, so that the concentration of pollutants is increased, and the abnormal increase B is judged.
Production abnormity: and data belonging to the characteristic production load are not matched, namely the exhaust emission of the enterprise is inconsistent with the power consumption of the production facility.
(3) Production anomaly A: the amount of exhaust emissions is abnormally low.
When the power consumption of the enterprise production facilities is increased and the exhaust emission is unchanged or reduced, the production is judged to be abnormal A.
Secondly, when the power consumption of the enterprise production facilities is unchanged and the exhaust emission is reduced, judging the production abnormality A.
(4) Production anomaly B: the amount of exhaust emissions is abnormally high.
When the power consumption of the enterprise production facilities is unchanged and the exhaust emission is increased, the enterprise production facilities are judged to be abnormal B.
And secondly, when the power consumption of the enterprise production facilities is reduced and the exhaust emission is increased or unchanged, judging that the production is abnormal B.
(5) Production stoppage abnormity:
and if parameters such as oxygen content, pollutant concentration, exhaust emission, electricity consumption of production facilities and the like in the production stopping process have any conditions which do not accord with normal production stopping conditions, judging that the production stopping is abnormal.
Firstly, abnormal production stop-abnormal oxygen content
And in the production stopping process, the oxygen content is unchanged or is lower than 19.5 percent, and the oxygen content is judged to be abnormal.
Production stop anomaly-contaminant concentration anomaly
And in the production stopping process, determining that the pollutant concentration is abnormal if the pollutant concentration is unchanged or higher than the normal production discharge.
Production stoppage abnormity-waste gas discharge abnormity
And in the production stopping process, judging that the exhaust emission is abnormal if the exhaust emission is unchanged or higher than the normal production emission.
Production stop abnormity-electricity consumption abnormity
And in the production stopping process, the electricity consumption of the production facility is unchanged or higher than the normal production emission, and the electricity consumption is judged to be abnormal.
In addition: the process of shutdown or startup can be carried out 5 hours after the shutdown and 5 hours before the shutdown, and the oxygen content, the exhaust emission and the pollutant concentration can fluctuate (each parameter is higher) in a reducing or increasing way, which belongs to the normal situation.
Although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the spirit and scope of the invention.
Claims (8)
1. An enterprise online monitoring data abnormity identification method is characterized by comprising the following steps:
1) acquiring enterprise online data and sub-meter electricity metering data, and preprocessing the data to obtain enterprise activity data;
2) generating a self-adaptive sliding window, and carrying out high-value detection on the enterprise activity data in the window to obtain a high-value sequence and a low-value sequence;
3) performing correlation analysis on the high and low value sequence data obtained in the step 2 to obtain correlation results among the data;
4) calling corresponding judgment logic according to the corresponding enterprise abnormal characteristics;
5) and judging the abnormal conditions and abnormal types of the enterprise activities.
2. The method for identifying abnormality in enterprise online monitoring data according to claim 1, wherein the step 1) includes:
(1) longitudinally processing the online data of the enterprise to obtain online pollutant parameters of the enterprise changing along with hours and online basic information parameters of the enterprise corresponding to the online pollutant parameters; longitudinally processing the sub-meter electricity metering data to obtain sub-meter electricity utilization parameters and corresponding sub-meter electricity metering basic information parameters;
(2) matching the processed online data of the enterprise with the sub-meter electricity metering data, and screening out the gas-related enterprises to be monitored and the corresponding information: the power consumption of the discharge port, the production facility and the processing facility;
(3) transversely processing the online data of the enterprise, screening out the row openings of the enterprise with serious online pollutant parameters loss and the corresponding discharge amount, and removing the row openings of the enterprise with insufficient data in continuous hours, so that enough data can enter the sliding window in the second step at the early stage;
(4) carrying out cubic spline interpolation on the cleaned enterprises and the data of the online main parameters of the corresponding row-opening enterprises, wherein the data is missing: a process of forming a smooth curve passing through a series of shape value points and obtaining a curve function set by solving a three bending moment equation set;
(5) and forming key parameter time series activity data corresponding to different gates of different enterprises and electric power time series activity data of production facilities and processing facilities matched with the gates.
3. The method of claim 2, wherein the online enterprise monitoring data anomaly identification method comprises SO2Monitoring concentration data, NOx monitoring concentration data and exhaust emission, wherein the enterprise online basic information parameters comprise: enterprise names, discharge port names, monitoring time, oxygen content and whether the operation is stopped or not, wherein the enterprise names, the discharge port names, the monitoring time and the oxygen content correspond to the online pollutant parameters of the enterprises; the sub-meter electricity consumption parameters comprise power of production facilities and processing facilities, and the sub-meter electricity consumption basic information parameters comprise enterprise names, facility types, monitoring point names and acquisition time.
4. The enterprise online monitoring data anomaly identification method according to claim 1, characterized by comprising the following steps: in the step 2), a self-adaptive sliding window method is adopted to identify a high value: aiming at different data types of different enterprises, combining the generated process characteristics and the data change rule expressed in daily production activities, adjusting the window size and the threshold variable setting, determining the effective identification characteristic high value, and establishing a characteristic sliding window aiming at different enterprise types.
5. The enterprise online monitoring data anomaly identification method according to claim 1, wherein the specific steps of the step 2) are as follows:
and (3) calculating the activity data in the step (1) in a sliding window: setting an initial sliding window size to be n, setting a signal vector signal to be 0,.. 0, setting a length to be an active data vector length L, setting an initial sliding window size variable n, a threshold variable threshold, a peak clipping variable influence, and setting an active data vector in a sliding window filter y to be y (1),... y (n), an arithmetic average in a current window needs to be calculated:
standard deviation within the current window:
the absolute value of the difference between the first data y (n + i) (i 1.. L-n) outside the current window and the mean of the active data within the current sliding window:
diff(n+i)＝|y(n+i)-avgFilter(n)| (3)
the fluctuation amount of the data vector in the window is as follows:
fluctuate(n)＝threshold*stdFilter(n) (4)
if diff (n + i) > fluctuate and y (n + i) > avgfilter (n) are satisfied, the signal value of the n + i-th data is set to 1 (peak signal), and if diff (n + i) > fluctuate and y (n + i) < avgfilter (n) at this time, the signal value of the n + i-th data is set to-1 (valley signal), the above process is followed by peak-clipping the filteredY (n + i) variable:
filteredY(n+i)＝influence*y(n+i)+(1-influence)*filteredY(n)(5)
if diff (n + i) < yield is satisfied, the signal value of the n + i-th data is set to 0 (wave-level signal), and filteredY (n + i) is assigned to y (n + i):
filteredY(n+i)＝y(n+i) (6)
and (3) cyclic calculation: and moving the whole window backward by one unit data, and calculating values (1) to (6) of a new window to obtain a signal value of the next data until the calculation of the data vector with the length of L is completed and a final signal vector signals is obtained.
6. The method for identifying the abnormality of the enterprise online monitoring data according to any one of claims 1-5, wherein the judging logic in the step 4) comprises:
through analyzing the online pollutant parameters of the enterprise of the online monitoring system of the enterprise and the change conditions of the sub-meter electricity utilization parameters of the sub-meter electricity system, combing the logical relationship among the parameters and the corresponding abnormal characteristics of the enterprise in the production and shutdown processes of the enterprise to obtain a judgment logic library of the abnormal characteristics of the enterprise, and calling the corresponding judgment logic according to the abnormal characteristics of the enterprise;
according to the production and shutdown states marked by the online monitoring system, enterprise activity data is divided into two types, namely production logic and shutdown logic, the production logic is called to judge in the production state, and the shutdown logic is called to judge in the shutdown state.
7. The enterprise online monitoring data anomaly identification method according to claim 6, characterized in that:
the production logic judging step is as follows:
the first step is as follows: judging whether the hour data change trends of the electricity consumption and the exhaust emission of the enterprise production facilities synchronously change or not: ascending/unchanging/reducing, if not synchronous, namely the changes of the waste gas emission quantity representing the production load and the facility power consumption are inconsistent, judging that the waste gas emission quantity and the facility power consumption do not accord with the production logic of the enterprise, and if synchronous, entering the second step;
the second step is that: judging whether the change trends of the power consumption of the processing facility and the exhaust emission are changed synchronously or not, if so, judging that the production and emission logic of an enterprise is met, and if not, judging that the production and emission logic is not met;
the third step: judging the abnormal emission type by analyzing the variation relation among the parameters, and entering the steps (1) - (4) of the step 5;
the production stop logic determination step comprises:
the first step is as follows: when the oxygen content is 19.5-22.0%, the pollutant concentration is far lower than the normal production emission: < average of historical production 20%, exhaust emissions of 0 or far below normal production emissions: < average historical production > 20%, electricity consumption for production is 0 or much lower than normal electricity consumption level: judging the production is normally stopped when the average historical production is 20 percent;
the second step is that: if any condition which does not accord with normal production halt appears in the production halt process, judging that the production halt is abnormal;
the third step: and (5) judging the type of the abnormal production halt by analyzing the data change conditions of the oxygen content, the pollutant concentration, the exhaust emission and the electricity consumption of the production facility, and entering the step 5.
8. The method for identifying abnormality of enterprise on-line monitoring data according to claim 6, characterized in that the abnormality type includes: abnormally rising A, abnormally rising B, abnormally producing A, abnormally producing B, abnormally stopping production;
wherein:
(1) abnormal rise A: the pollutant concentration is not in accordance with the logic of the exhaust emission, the power consumption of production facilities and the power consumption of treatment facilities,
when the data of the waste gas emission amount of an enterprise and the power consumption of a production facility are not changed, the power consumption of a processing facility is increased, the pollutant concentration is increased, and the abnormal increase A is judged,
when the data of the waste gas emission of the enterprise and the power consumption of the production facility are reduced, the power consumption of the processing facility is unchanged, and the concentration of pollutants is increased, judging that the concentration is abnormally increased A;
(2) abnormal rise B: the handling measures do not match the production load,
when the exhaust emission of an enterprise is increased, the power consumption of production facilities is increased, but the power consumption of treatment facilities is unchanged or is reduced to cause the increase of pollutant concentration, the abnormal increase B is judged,
when the exhaust emission of an enterprise and the power consumption of a production facility are unchanged, the power consumption of a treatment facility is reduced, so that the concentration of pollutants is increased, and the concentration is judged to be abnormally increased B;
(3) production anomaly A: the amount of exhaust emissions is extremely low,
firstly, when the power consumption of the enterprise production facilities is increased and the exhaust emission is unchanged or reduced, the enterprise production facilities are judged to be abnormal A,
secondly, when the power consumption of the enterprise production facilities is unchanged and the exhaust emission is reduced, judging the enterprise production facilities to be abnormal A;
(4) production anomaly B: the amount of exhaust emissions is abnormally high,
firstly, when the power consumption of the enterprise production facilities is not changed and the exhaust emission is increased, the enterprise production facilities are judged to be abnormal B,
judging that the production is abnormal B when the power consumption of the enterprise production facilities is reduced and the exhaust emission is increased or unchanged;
(5) production stoppage abnormity:
firstly, abnormal production stop-abnormal oxygen content
In the production stopping process, the oxygen content is not changed or is lower than 19.5 percent, the oxygen content is judged to be abnormal,
production stop anomaly-contaminant concentration anomaly
In the process of stopping production, the concentration of the pollutants is not changed or is higher than the normal production discharge, the concentration of the pollutants is judged to be abnormal,
production stoppage abnormity-waste gas discharge abnormity
In the process of stopping production, the exhaust emission is not changed or is higher than the normal production emission, the exhaust emission is judged to be abnormal,
production stop abnormity-electricity consumption abnormity
And in the production stopping process, the electricity consumption of the production facility is unchanged or higher than the normal production emission, and the electricity consumption is judged to be abnormal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110974638.5A CN113792988B (en) | 2021-08-24 | 2021-08-24 | Enterprise online monitoring data anomaly identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110974638.5A CN113792988B (en) | 2021-08-24 | 2021-08-24 | Enterprise online monitoring data anomaly identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113792988A true CN113792988A (en) | 2021-12-14 |
| CN113792988B CN113792988B (en) | 2024-07-12 |
Family
ID=78876360
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110974638.5A Active CN113792988B (en) | 2021-08-24 | 2021-08-24 | Enterprise online monitoring data anomaly identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113792988B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114693166A (en) * | 2022-04-22 | 2022-07-01 | 北京中科三清环境技术有限公司 | Method and device for monitoring execution condition of emergency emission reduction measure on line and electronic equipment |
| CN115238753A (en) * | 2022-09-21 | 2022-10-25 | 西南交通大学 | Self-adaptive SHM data cleaning method based on local outlier factor |
| CN115689396A (en) * | 2022-12-30 | 2023-02-03 | 天津友美环保科技有限公司 | Pollutant discharge control method, device, equipment and medium |
| CN117237165A (en) * | 2023-09-22 | 2023-12-15 | 中节能天融科技有限公司 | Method for detecting fake data |
| CN117609928A (en) * | 2024-01-23 | 2024-02-27 | 华北电力科学研究院有限责任公司 | Pollutant emission amount anomaly identification method and device based on power data |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101478776A (en) * | 2009-01-23 | 2009-07-08 | 北京交通大学 | Method of railway wireless environment path loss index estimation |
| CN104571050A (en) * | 2015-01-13 | 2015-04-29 | 国网上海市电力公司 | Industrial enterprise pollutant emission monitoring method based on high-frequency power data |
| US20170371757A1 (en) * | 2016-06-28 | 2017-12-28 | Beijing Baidu Netcom Science And Technology, Ltd. | System monitoring method and apparatus |
| CN108876181A (en) * | 2018-06-30 | 2018-11-23 | 珠海市万顺睿通科技有限公司 | Intelligentized enterprise pollution object exhausting control method and system |
| CN110008565A (en) * | 2019-03-28 | 2019-07-12 | 浙江大学 | A kind of industrial process unusual service condition prediction technique based on operating parameter association analysis |
| CN110414045A (en) * | 2019-06-18 | 2019-11-05 | 东华大学 | Short-term wind speed forecasting method based on VMD-GRU |
| WO2020010701A1 (en) * | 2018-07-11 | 2020-01-16 | 平安科技（深圳）有限公司 | Pollutant anomaly monitoring method and system, computer device, and storage medium |
| CN110889088A (en) * | 2019-11-04 | 2020-03-17 | 国网浙江省电力有限公司信息通信分公司 | Enterprise pollution discharge supervision method assisted by electric model |
| CN111191881A (en) * | 2019-12-13 | 2020-05-22 | 大唐东北电力试验研究院有限公司 | Thermal power generating unit industrial equipment state monitoring method based on big data |
| CN111275307A (en) * | 2020-01-16 | 2020-06-12 | 生态环境部华南环境科学研究所 | Quality control method for high-frequency continuous observation data of automatic online water quality station |
| CN211628078U (en) * | 2020-04-15 | 2020-10-02 | 江苏安科瑞电器制造有限公司 | Environmental control process monitoring system |
| CN112101969A (en) * | 2020-09-27 | 2020-12-18 | 西安交通大学 | Environmental protection data false-making detection method based on time sequence sliding window discrete coefficient |
| CN112508267A (en) * | 2020-12-04 | 2021-03-16 | 国网湖南省电力有限公司 | Distribution transformer area load prediction method and system based on artificial intelligence |
| CN112862605A (en) * | 2021-04-26 | 2021-05-28 | 广东浩迪智云技术有限公司 | Enterprise operation abnormity prompting method and system based on electricity consumption data |
-
2021
- 2021-08-24 CN CN202110974638.5A patent/CN113792988B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101478776A (en) * | 2009-01-23 | 2009-07-08 | 北京交通大学 | Method of railway wireless environment path loss index estimation |
| CN104571050A (en) * | 2015-01-13 | 2015-04-29 | 国网上海市电力公司 | Industrial enterprise pollutant emission monitoring method based on high-frequency power data |
| US20170371757A1 (en) * | 2016-06-28 | 2017-12-28 | Beijing Baidu Netcom Science And Technology, Ltd. | System monitoring method and apparatus |
| CN108876181A (en) * | 2018-06-30 | 2018-11-23 | 珠海市万顺睿通科技有限公司 | Intelligentized enterprise pollution object exhausting control method and system |
| WO2020010701A1 (en) * | 2018-07-11 | 2020-01-16 | 平安科技（深圳）有限公司 | Pollutant anomaly monitoring method and system, computer device, and storage medium |
| CN110008565A (en) * | 2019-03-28 | 2019-07-12 | 浙江大学 | A kind of industrial process unusual service condition prediction technique based on operating parameter association analysis |
| CN110414045A (en) * | 2019-06-18 | 2019-11-05 | 东华大学 | Short-term wind speed forecasting method based on VMD-GRU |
| CN110889088A (en) * | 2019-11-04 | 2020-03-17 | 国网浙江省电力有限公司信息通信分公司 | Enterprise pollution discharge supervision method assisted by electric model |
| CN111191881A (en) * | 2019-12-13 | 2020-05-22 | 大唐东北电力试验研究院有限公司 | Thermal power generating unit industrial equipment state monitoring method based on big data |
| CN111275307A (en) * | 2020-01-16 | 2020-06-12 | 生态环境部华南环境科学研究所 | Quality control method for high-frequency continuous observation data of automatic online water quality station |
| CN211628078U (en) * | 2020-04-15 | 2020-10-02 | 江苏安科瑞电器制造有限公司 | Environmental control process monitoring system |
| CN112101969A (en) * | 2020-09-27 | 2020-12-18 | 西安交通大学 | Environmental protection data false-making detection method based on time sequence sliding window discrete coefficient |
| CN112508267A (en) * | 2020-12-04 | 2021-03-16 | 国网湖南省电力有限公司 | Distribution transformer area load prediction method and system based on artificial intelligence |
| CN112862605A (en) * | 2021-04-26 | 2021-05-28 | 广东浩迪智云技术有限公司 | Enterprise operation abnormity prompting method and system based on electricity consumption data |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114693166A (en) * | 2022-04-22 | 2022-07-01 | 北京中科三清环境技术有限公司 | Method and device for monitoring execution condition of emergency emission reduction measure on line and electronic equipment |
| CN115238753A (en) * | 2022-09-21 | 2022-10-25 | 西南交通大学 | Self-adaptive SHM data cleaning method based on local outlier factor |
| CN115238753B (en) * | 2022-09-21 | 2022-12-06 | 西南交通大学 | Self-adaptive SHM data cleaning method based on local outlier factor |
| CN115689396A (en) * | 2022-12-30 | 2023-02-03 | 天津友美环保科技有限公司 | Pollutant discharge control method, device, equipment and medium |
| CN115689396B (en) * | 2022-12-30 | 2023-08-08 | 天津友美环保科技有限公司 | Pollutant emission control method, device, equipment and medium |
| CN117237165A (en) * | 2023-09-22 | 2023-12-15 | 中节能天融科技有限公司 | Method for detecting fake data |
| CN117237165B (en) * | 2023-09-22 | 2024-05-07 | 中节能数字科技有限公司 | Method for detecting fake data |
| CN117609928A (en) * | 2024-01-23 | 2024-02-27 | 华北电力科学研究院有限责任公司 | Pollutant emission amount anomaly identification method and device based on power data |
| CN117609928B (en) * | 2024-01-23 | 2024-04-05 | 华北电力科学研究院有限责任公司 | Pollutant emission amount anomaly identification method and device based on power data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113792988B (en) | 2024-07-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113792988A (en) | Online monitoring data anomaly identification method for enterprise | |
| CN110208019B (en) | Dynamic threshold early warning method for monitoring state of mobile equipment | |
| CN110895526A (en) | Method for correcting data abnormity in atmosphere monitoring system | |
| CN110889088A (en) | Enterprise pollution discharge supervision method assisted by electric model | |
| CN114757380B (en) | Fault early warning system and method for thermal power plant, electronic equipment and storage medium | |
| CN112465239B (en) | Desulfurization system operation optimization method based on improved PSO-FCM algorithm | |
| CN110094251B (en) | SCR catalyst performance degradation analysis method based on time-interval multi-model modeling | |
| CN112287988A (en) | Method for identifying water pollution source online monitoring data abnormity | |
| CN113988711B (en) | Pollution discharge enterprise stop or limit production monitoring method based on electricity consumption data in control state | |
| CN116151621A (en) | Atmospheric pollution treatment risk detection system based on data analysis | |
| CN112001511A (en) | Equipment reliability and dynamic risk evaluation method, system and equipment based on data mining | |
| CN111289697A (en) | CEMS running state monitoring and data false identification method and system | |
| CN114757413A (en) | Bad data identification method based on time sequence series analysis coupling neural network prediction | |
| CN110956268A (en) | Compressor fault diagnosis method based on big data | |
| CN112037106B (en) | Data anomaly analysis method based on feature cross-correlation and probability density | |
| CN116363843A (en) | Laboratory equipment early warning system | |
| CN117034197A (en) | Enterprise power consumption typical mode analysis method based on multidimensional Isolate-detection multi-point detection | |
| CN116805065A (en) | Intelligent management method for monitoring data of coal powder heating furnace burner | |
| CN115879726A (en) | Method for mutually optimizing and screening power consumption and emission data of enterprise | |
| CN112580838B (en) | Environment protection equipment operation and maintenance management method, system, computer storage medium and terminal | |
| CN115115213A (en) | Method and device for retrospective analysis of quality characteristics of steel products | |
| CN111624985B (en) | Gas turbine control system sensor fault diagnosis method | |
| CN112016050B (en) | CEMS system abnormal data monitoring method based on manifold learning | |
| CN115062851A (en) | Pollution discharge abnormity monitoring method and system based on multi-algorithm fusion | |
| CN113986892A (en) | Coal consumption prediction and data cleaning method and system for thermal power generating unit |
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 | ||
| GR01 | Patent grant |