Fuel monitoring system and fuel monitoring method

Abstract

The present invention provides a fuel monitoring system and a monitoring method, which relate to the technical field of fuel monitoring, and include a data acquisition module: used to use a target sensor to collect first characteristic data of a target fuel in real time; a data processing module: used to preprocess the first characteristic data to obtain target characteristic data; a monitoring module: used to realize abnormal state detection and prediction of the target fuel based on the analysis of the target characteristic data, and generate a corresponding fuel monitoring report; an alarm module: used to realize state alarm and early warning functions of the target fuel based on the fuel monitoring report. By using different sensors to collect the characteristic data of the target fuel in real time and analyze it, abnormal state detection and prediction of the target fuel are realized, and the state alarm and early warning functions of the target fuel are realized through the generated fuel monitoring report, which can realize effective supervision of fuel quality and ensure fuel production safety and high efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of fuel monitoring, and in particular to a fuel monitoring system and a monitoring method. Background Art

[0002] In recent years, fuel has played an important and indispensable role in industrial and agricultural production, transportation, and people's lives, bringing huge benefits and convenience to society. However, the instability of fuel status has brought a series of problems to the industry, such as unstable combustion effect, high energy consumption, environmental pollution, etc. Traditional fuel status monitoring systems or methods have problems such as low fuel status detection accuracy and untimely detection of safety hazards. Therefore, how to effectively monitor the fuel status to ensure the safety and efficiency of resource production has become one of the current research and development centers.

[0003] Therefore, the present invention provides a fuel monitoring system and a monitoring method. Summary of the invention

[0004] The present invention provides a fuel monitoring system and a monitoring method, which are used to collect and analyze the characteristic data of the target fuel in real time by using different sensors, realize the detection and prediction of the abnormal state of the target fuel, and realize the state alarm and early warning function of the target fuel through the generated fuel monitoring report, so as to realize the effective supervision of the fuel quality and ensure the safety of fuel production.

[0005] The present invention provides a fuel monitoring system, comprising:

[0006] Data acquisition module: used for acquiring first characteristic data of target fuel in real time by using target sensor;

[0007] Data processing module: used for preprocessing the first characteristic data to obtain target characteristic data;

[0008] Monitoring module: used to detect and predict the abnormal state of the target fuel based on the analysis of the target characteristic data, and generate a corresponding fuel monitoring report;

[0009] Alarm module: used to realize the status alarm and early warning function of the target fuel based on the fuel monitoring report.

[0010] Preferably, the first characteristic data includes sulfur content, ash content and calorific value.

[0011] Preferably, the monitoring module includes:

[0012] Annotation unit: used to compare the target characteristic data with the set characteristic threshold. If the target characteristic data exceeds the set characteristic threshold and the characteristic difference does not belong to the preset characteristic range, the current target characteristic data is determined to be abnormal, and the current target characteristic data is marked as a first-level abnormality;

[0013] If the target characteristic data exceeds the set characteristic threshold and the characteristic difference falls within the preset characteristic range, the current target characteristic data is determined to be abnormal, and the current target characteristic data is marked as a secondary abnormality;

[0014] If the target characteristic data does not exceed the set characteristic threshold, the current target characteristic data is marked as normal;

[0015] Status Analysis Unit:

[0016] Level judgment block: used to judge that the current target fuel state is normal if all target characteristic data of the target fuel at the current moment are normal;

[0017] Otherwise, the abnormality assessment index is calculated based on the abnormality level of the target characteristic data at the current moment;

[0018] The calculation formula of the abnormal assessment index is as follows:

[0019] ;

[0020] Where y represents the abnormal evaluation index of the target fuel with abnormal target characteristic data; It is indicated as setting the first-level abnormal risk factor; It represents the number of target characteristic data with first-level anomalies; X represents the total number of target characteristic data; It is represented as the characteristic difference between the i-th target characteristic data marked with the first-level anomaly and the set characteristic threshold; It is represented as the influence weight of the i-th target characteristic data with the first-level anomaly marked on the fuel anomaly assessment; It is expressed as setting the secondary abnormal risk factor; It is expressed as the number of target characteristic data with marked secondary anomalies; It is represented as the characteristic difference between the i-th target characteristic data marked as the secondary anomaly and the set characteristic threshold; It is represented as the influence weight of the i-th target characteristic data that marks the secondary anomaly on the fuel anomaly assessment;

[0021] Based on the set abnormal range table, according to the abnormal evaluation index, the abnormal level of the current target fuel is determined;

[0022] Prediction block: used to predict the state of the target fuel at a preset future moment when the target fuel state is normal, and generate a state abnormality prediction result;

[0023] Report generation unit: used to summarize the abnormal level, abnormal evaluation index, target characteristic data, characteristic difference and basic fuel information of the target fuel when the target fuel state is abnormal, and generate a fuel monitoring report;

[0024] When the target fuel state is normal, the target characteristic data, characteristic difference, abnormal state prediction results and basic fuel information of the target fuel are summarized to generate a fuel monitoring report.

[0025] Preferably, the prediction block includes:

[0026] Model building sub-block: used to extract historical data collected by target sensors in different historical time periods from the equipment operation database to expand the data to obtain a training data set;

[0027] Extracting the data time series features in the training data set to construct a first prediction training data feature set, and training a neural network to obtain a data prediction model;

[0028] Extracting abnormal features from the training data set to construct a second prediction training data feature set, training a neural network, and obtaining an abnormal prediction model;

[0029] Prediction sub-block: used to input the target characteristic data when the target fuel state is normal into the data prediction model, and output the possible characteristic value at a preset future moment;

[0030] Anomaly detection sub-block: used to use the anomaly prediction model and the labeling unit to make two anomaly predictions on the target characteristic data state at the preset future moment based on the possible characteristic values ​​at the preset future moment, and analyze the two anomaly prediction results to obtain the anomaly prediction result of the state of the target fuel at the preset future moment.

[0031] Preferably, the anomaly detection sub-block includes:

[0032] Inputting target characteristic data when the target fuel state is normal into the abnormality prediction model, and outputting a first abnormality prediction result of the target characteristic data at a preset future moment;

[0033] Performing anomaly detection on possible characteristic values ​​at a preset future time using a labeling unit to obtain a second anomaly prediction result corresponding to the target characteristic data;

[0034] Compare the state determination result in the first abnormality prediction result with the state determination result in the second abnormality prediction result, and if the state determination results of the current target characteristic data are consistent, use the second abnormality prediction result as a reference abnormality prediction result of the current target characteristic data;

[0035] If the state determination result of the current target characteristic data is inconsistent, the possible characteristic value of the current target characteristic data at the preset future time is adjusted to obtain a corrected characteristic value;

[0036] Performing anomaly detection on the modified characteristic value by using a labeling unit to obtain a third anomaly prediction result corresponding to the target characteristic data, and using the third anomaly prediction result as a reference anomaly prediction result for the current target characteristic data;

[0037] Using the state analysis unit to preset a reference abnormality prediction result at a future moment based on all target characteristic data of the target fuel, perform future state analysis on the fuel to obtain a first state prediction result;

[0038] Determine the abnormal risk index of the target fuel at a preset future time based on the set risk factor of the abnormal level of the target characteristic data;

[0039] The first state prediction result is combined with the abnormal risk index and output as the state abnormality prediction result.

[0040] Preferably, the calculation formula for the modified characteristic value is as follows:

[0041] ;

[0042] In the formula, It is expressed as the corrected characteristic value of the ath target characteristic data; It is represented as the possible characteristic value of the ath target characteristic data; It is represented as the characteristic value of the a-th target characteristic data at the current moment; Expressed as data prediction error compensation factor; It is expressed as the average of the absolute differences between the characteristic value of the a-th target characteristic data at the current moment and the characteristic values ​​at the previous three moments; It is expressed as the average of the absolute differences between the possible characteristic values ​​of the a-th target characteristic data and the characteristic values ​​at the current moment and the two moments before the current moment; It represents the data acquisition loss factor of the target sensor; Expressed as a data prediction bias compensation factor.

[0043] Preferably, the alarm module comprises:

[0044] Transmission unit: used for transmitting the corresponding fuel monitoring report to the alarm unit according to the target fuel state in the received fuel monitoring report, if the target fuel state is abnormal;

[0045] If the target fuel status is normal, the corresponding fuel monitoring report is transmitted to the early warning unit;

[0046] Alarm unit: used to extract the list of maintenance personnel for the target fuel from the fuel maintenance database according to the basic fuel information of the target fuel in the received fuel monitoring report;

[0047] Filtering a target fuel status maintenance personnel who matches the abnormality level of the target fuel from the maintenance processing personnel list;

[0048] Transmit the fuel monitoring report to the fuel status maintenance personnel to maintain the target fuel status;

[0049] Early warning unit: used to generate a target fuel early warning task according to the received fuel monitoring report, if the target fuel state is abnormal at a preset future time, and add the target fuel early warning task to the early warning processing task column according to the abnormal risk index;

[0050] The warning processing personnel maintain the fuel status according to the warning processing task list.

[0051] The present invention provides a fuel monitoring method, comprising:

[0052] Step 1: using a target sensor to collect first characteristic data of a target fuel in real time;

[0053] Step 2: preprocessing the first characteristic data to obtain target characteristic data;

[0054] Step 3: Based on the analysis of the target characteristic data, abnormal state detection and prediction of the target fuel are achieved, and a corresponding fuel monitoring report is generated;

[0055] Step 4: Based on the fuel monitoring report, realize the status alarm and early warning function of the target fuel.

[0056] Compared with the prior art, the beneficial effects of this application are as follows:

[0057] By using different sensors to collect and analyze the characteristic data of the target fuel in real time, the abnormal state detection and prediction of the target fuel can be achieved, and the state alarm and early warning function of the target fuel can be realized through the generated fuel monitoring report, which can effectively supervise the fuel quality and ensure the safety and high efficiency of fuel production.

[0058] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0059] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0061] Figure 1 A structural diagram of a fuel monitoring system in an embodiment of the present invention;

[0062] Figure 2 The figure is a flow chart of a fuel monitoring method in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0064] An embodiment of the present invention provides a fuel monitoring system, such as Figure 1 As shown, including:

[0065] Data acquisition module: used for acquiring first characteristic data of target fuel in real time by using target sensor;

[0066] Data processing module: used for preprocessing the first characteristic data to obtain target characteristic data;

[0067] Monitoring module: used to detect and predict the abnormal state of the target fuel based on the analysis of the target characteristic data, and generate a corresponding fuel monitoring report;

[0068] Alarm module: used to realize the status alarm and early warning function of the target fuel based on the fuel monitoring report.

[0069] In this embodiment, the target sensor is used to collect first characteristic data of the target fuel in real time, wherein the first characteristic data includes sulfur content, ash content and calorific value; the target characteristic data is data obtained after cleaning the first characteristic data.

[0070] In this embodiment, the fuel monitoring report includes two types, one of which is a monitoring report generated by summarizing the abnormal level, abnormal evaluation index, target characteristic data, characteristic difference and fuel basic information of the target fuel when the target fuel state is abnormal, wherein the abnormal evaluation index is used to characterize the abnormal degree of the target fuel with abnormal target characteristic data, the characteristic difference refers to the difference obtained by comparing the target characteristic data with the set characteristic threshold, and the set characteristic threshold is set in advance; the fuel basic information includes the fuel name, fuel type and fuel location;

[0071] The other is a monitoring report generated by summarizing the target characteristic data, characteristic difference, abnormal state prediction results and basic fuel information of the target fuel when the target fuel state is normal, wherein the abnormal state prediction result refers to the target fuel state result at a preset future moment predicted based on the target characteristic data of the normal current state of the target fuel, including state judgment results, abnormal risk index, fuel abnormality level, etc.

[0072] The beneficial effects of the above technical solution are: by using different sensors to collect and analyze the characteristic data of the target fuel in real time, the abnormal state detection and prediction of the target fuel can be realized, and the state alarm and early warning function of the target fuel can be realized through the generated fuel monitoring report, which can realize effective supervision of fuel quality and ensure the safety and high efficiency of fuel production.

[0073] An embodiment of the present invention provides a fuel monitoring system, wherein the monitoring module comprises:

[0074] Annotation unit: used to compare the target characteristic data with the set characteristic threshold. If the target characteristic data exceeds the set characteristic threshold and the characteristic difference does not belong to the preset characteristic range, the current target characteristic data is determined to be abnormal, and the current target characteristic data is marked as a first-level abnormality;

[0075] If the target characteristic data exceeds the set characteristic threshold and the characteristic difference falls within the preset characteristic range, the current target characteristic data is determined to be abnormal, and the current target characteristic data is marked as a secondary abnormality;

[0076] If the target characteristic data does not exceed the set characteristic threshold, the current target characteristic data is marked as normal;

[0077] Status Analysis Unit:

[0078] Level judgment block: used to judge that the current target fuel state is normal if all target characteristic data of the target fuel at the current moment are normal;

[0079] Otherwise, the abnormality assessment index is calculated based on the abnormality level of the target characteristic data at the current moment;

[0080] The calculation formula of the abnormal assessment index is as follows:

[0081] ;

[0082] Where y represents the abnormal evaluation index of the target fuel with abnormal target characteristic data; It is indicated as setting the first-level abnormal risk factor; It represents the number of target characteristic data with first-level anomalies; X represents the total number of target characteristic data; It is represented as the characteristic difference between the i-th target characteristic data marked with the first-level anomaly and the set characteristic threshold; It is represented as the influence weight of the i-th target characteristic data with the first-level anomaly marked on the fuel anomaly assessment; It is expressed as setting the secondary abnormal risk factor; It is expressed as the number of target characteristic data with marked secondary anomalies; It is represented as the characteristic difference between the i-th target characteristic data marked as the secondary anomaly and the set characteristic threshold; It is represented as the influence weight of the i-th target characteristic data that marks the secondary anomaly on the fuel anomaly assessment;

[0083] Based on the set abnormal range table, according to the abnormal evaluation index, the abnormal level of the current target fuel is determined;

[0084] Prediction block: used to predict the state of the target fuel at a preset future moment when the target fuel state is normal, and generate a state abnormality prediction result;

[0085] Report generation unit: used to summarize the abnormal level, abnormal evaluation index, target characteristic data, characteristic difference and basic fuel information of the target fuel when the target fuel state is abnormal, and generate a fuel monitoring report;

[0086] When the target fuel state is normal, the target characteristic data, characteristic difference, abnormal state prediction results and basic fuel information of the target fuel are summarized to generate a fuel monitoring report.

[0087] In this embodiment, the target characteristic data is the data obtained after cleaning the first characteristic data, wherein the first characteristic data includes sulfur content, ash content and calorific value; the set characteristic threshold is set in advance; the characteristic difference refers to the difference obtained by comparing the target characteristic data with the set characteristic threshold, and the preset characteristic range is determined in advance.

[0088] In this embodiment, the abnormality level of the target characteristic data includes two types: primary abnormality and secondary abnormality.

[0089] In this embodiment, for example, there are target characteristic data g1, g2, and g3, where g1 and g2 exceed the set characteristic threshold, and the characteristic difference with the set characteristic threshold belongs to the preset characteristic range; g3 does not exceed the set characteristic threshold. At this time, the target characteristic data g1 and g2 are respectively marked as secondary abnormalities, and the target characteristic data g3 is marked as normal.

[0090] In this embodiment, the abnormality assessment index is used to characterize the abnormality degree of the target fuel with abnormal target characteristic data; the abnormal range table is set in advance and is used to determine the abnormality level of the target fuel, wherein the abnormality level of the target fuel includes level one and level two; the preset future moment refers to the next moment; the state abnormality prediction result refers to the target fuel state result at the preset future moment obtained based on the target characteristic data of the normal current state of the target fuel, including the state judgment result, abnormal risk index, fuel abnormality level, etc., wherein the state judgment result includes normal and abnormal.

[0091] The beneficial effect of the above technical solution is: by monitoring the state of the target characteristic data of the target fuel and summarizing and analyzing the state results of all target characteristic data, the state detection and prediction of the target fuel can be achieved, which can effectively determine the fuel safety hazards and thus ensure the production safety of the target fuel.

[0092] An embodiment of the present invention provides a fuel monitoring system, wherein the prediction block comprises:

[0093] Model building sub-block: used to extract historical data collected by target sensors in different historical time periods from the equipment operation database to expand the data to obtain a training data set;

[0094] Extracting the data time series features in the training data set to construct a first prediction training data feature set, and training a neural network to obtain a data prediction model;

[0095] Extracting abnormal features from the training data set to construct a second prediction training data feature set, training a neural network, and obtaining an abnormal prediction model;

[0096] Prediction sub-block: used to input the target characteristic data when the target fuel state is normal into the data prediction model, and output the possible characteristic value at a preset future moment;

[0097] Anomaly detection sub-block: used to use the anomaly prediction model and the labeling unit to make two anomaly predictions on the target characteristic data state at the preset future moment based on the possible characteristic values ​​at the preset future moment, and analyze the two anomaly prediction results to obtain the anomaly prediction result of the state of the target fuel at the preset future moment.

[0098] In this embodiment, the equipment operation database is composed of all equipment involved in the fuel production process and equipment operation data and equipment collection data, such as coal-fired furnaces and sensors; the training data set is a data set obtained by expanding the historical data collected by the target sensor in different historical time periods.

[0099] In this embodiment, the first prediction training data feature set is constructed based on the data time series features extracted from the training data set; the data prediction model is a prediction model obtained by training the neural network using the first prediction training data feature set, and is used to predict the numerical value of the target characteristic data at future moments; the second prediction training data feature set is constructed based on the abnormal features extracted from the training data set; the abnormal prediction model is a model obtained by training the neural network using the second prediction training data feature set, and is used to predict the abnormal results of the target characteristic data at future moments.

[0100] In this embodiment, the possible characteristic value is the value obtained by inputting the target characteristic data when the target fuel state is normal into the data prediction model; the abnormal state prediction result refers to the target fuel state result at a preset future moment obtained by predicting the target characteristic data when the target fuel is currently in a normal state, including state judgment results, abnormal risk index, fuel abnormality level, etc., among which the state judgment results include normal and abnormal.

[0101] The beneficial effect of the above technical solution is: by constructing a data prediction model and an abnormal prediction model to perform abnormal prediction analysis on the target characteristic data of the target fuel at a preset future time, and then accurately predicting the state of the target fuel at the future time, fuel safety hazards can be effectively avoided, which is conducive to ensuring fuel production safety.

[0102] An embodiment of the present invention provides a fuel monitoring system, wherein the abnormality detection sub-block includes:

[0103] Inputting target characteristic data when the target fuel state is normal into the abnormality prediction model, and outputting a first abnormality prediction result of the target characteristic data at a preset future moment;

[0104] Performing anomaly detection on possible characteristic values ​​at a preset future time using a labeling unit to obtain a second anomaly prediction result corresponding to the target characteristic data;

[0105] Compare the state determination result in the first abnormality prediction result with the state determination result in the second abnormality prediction result, and if the state determination results of the current target characteristic data are consistent, use the second abnormality prediction result as a reference abnormality prediction result of the current target characteristic data;

[0106] If the state determination result of the current target characteristic data is inconsistent, the possible characteristic value of the current target characteristic data at the preset future time is adjusted to obtain a corrected characteristic value;

[0107] Performing anomaly detection on the modified characteristic value by using a labeling unit to obtain a third anomaly prediction result corresponding to the target characteristic data, and using the third anomaly prediction result as a reference anomaly prediction result for the current target characteristic data;

[0108] Using the state analysis unit to preset a reference abnormality prediction result at a future moment based on all target characteristic data of the target fuel, perform future state analysis on the fuel to obtain a first state prediction result;

[0109] Determine the abnormal risk index of the target fuel at a preset future time based on the set risk factor of the abnormal level of the target characteristic data;

[0110] The first state prediction result is combined with the abnormal risk index and output as the state abnormality prediction result.

[0111] In this embodiment, the target characteristic data is the data obtained after cleaning the first characteristic data, wherein the first characteristic data includes sulfur content, ash content and calorific value; the abnormal prediction model is a model obtained by training a neural network using a second prediction training data feature set, and is used to predict abnormal results of the target characteristic data at future moments, wherein the second prediction training data feature set is constructed based on abnormal features extracted from the training data set; the training data set is a data set obtained by expanding historical data collected by the target sensor in different historical time periods.

[0112] In this embodiment, the first abnormal prediction result is used to represent the state judgment result of the target characteristic data, wherein the state judgment result includes normal and abnormal; the second abnormal prediction result is a prediction result obtained by performing abnormal detection on the possible characteristic value using the marking unit, including the state judgment result and the abnormal level; the corrected characteristic value is a value obtained by correcting the possible characteristic value; the third abnormal prediction result is a result obtained by performing abnormal detection on the corrected characteristic value using the marking unit; the first state prediction result refers to the state prediction of the target fuel at a preset future moment.

[0113] In this embodiment, the set risk factor is set in advance; the abnormal risk index is calculated based on the number of target characteristic data of different abnormal levels combined with the set risk factor of the corresponding abnormal level, and is used to express the abnormal threat level of the target fuel at a preset future moment.

[0114] The beneficial effect of the above technical solution is: by comparing the abnormal prediction results obtained by twice predicting the target characteristic data when the target fuel state is normal using the labeling unit and the abnormal prediction model, the abnormal prediction results of the target characteristic data are summarized and analyzed to obtain the abnormal prediction results of the state of the target fuel, which is helpful to avoid fuel safety hazards and achieve effective supervision of the fuel.

[0115] An embodiment of the present invention provides a fuel monitoring system, and the calculation formula of the correction characteristic value is as follows:

[0116] ;

[0117] In the formula, It is expressed as the corrected characteristic value of the ath target characteristic data; It is represented as the possible characteristic value of the ath target characteristic data; It is represented as the characteristic value of the a-th target characteristic data at the current moment; Expressed as data prediction error compensation factor; It is expressed as the average of the absolute differences between the characteristic value of the a-th target characteristic data at the current moment and the characteristic values ​​at the previous three moments; It is expressed as the average of the absolute differences between the possible characteristic values ​​of the a-th target characteristic data and the characteristic values ​​at the current moment and the two moments before the current moment; It represents the data acquisition loss factor of the target sensor; Expressed as a data prediction bias compensation factor.

[0118] The beneficial effect of the above technical solution is: by correcting the possible characteristic values ​​to obtain the corrected characteristic values, an accurate data basis is provided for the abnormal state prediction of the target characteristic data, thereby ensuring the accurate state prediction of the target fuel.

[0119] An embodiment of the present invention provides a fuel monitoring system, wherein the alarm module comprises:

[0120] Transmission unit: used for transmitting the corresponding fuel monitoring report to the alarm unit according to the target fuel state in the received fuel monitoring report, if the target fuel state is abnormal;

[0121] If the target fuel status is normal, the corresponding fuel monitoring report is transmitted to the early warning unit;

[0122] Alarm unit: used to extract the list of maintenance personnel for the target fuel from the fuel maintenance database according to the basic fuel information of the target fuel in the received fuel monitoring report;

[0123] Filtering a target fuel status maintenance personnel who matches the abnormality level of the target fuel from the maintenance processing personnel list;

[0124] Transmit the fuel monitoring report to the fuel status maintenance personnel to maintain the target fuel status;

[0125] Early warning unit: used to generate a target fuel early warning task according to the received fuel monitoring report, if the target fuel state is abnormal at a preset future time, and add the target fuel early warning task to the early warning processing task column according to the abnormal risk index;

[0126] The warning processing personnel maintain the fuel status according to the warning processing task list.

[0127] In this embodiment, the target fuel status includes normal and abnormal; the basic fuel information includes the fuel name, fuel type and fuel location; the fuel maintenance database is a database consisting of the basic fuel information and fuel maintenance personnel; the maintenance personnel list is a list of maintenance personnel corresponding to the target fuel.

[0128] In this embodiment, the target fuel status maintenance personnel refers to the maintenance processing personnel who match the abnormal level of the target fuel; the abnormal risk index is calculated based on the number of target characteristic data of different abnormal levels combined with the set risk coefficient of the corresponding abnormal level, and is used to express the abnormal threat level of the target fuel at a preset future moment; the early warning processing task column is composed of a number of fuel early warning tasks in descending order of the abnormal risk index; the early warning processing personnel is used to process the early warning processing tasks to respond to the fuel early warning.

[0129] The beneficial effect of the above technical solution is: by receiving and analyzing the fuel status detection report to provide status alarms and warnings for the target fuel, safe supervision of the fuel status is effectively achieved, fuel production safety is ensured, and fuel production efficiency is improved.

[0130] The embodiment of the present invention provides a fuel monitoring method, such as Figure 2 As shown, including:

[0131] Step 1: using a target sensor to collect first characteristic data of a target fuel in real time;

[0132] Step 2: preprocessing the first characteristic data to obtain target characteristic data;

[0133] Step 3: Based on the analysis of the target characteristic data, abnormal state detection and prediction of the target fuel are achieved, and a corresponding fuel monitoring report is generated;

[0134] Step 4: Based on the fuel monitoring report, realize the status alarm and early warning function of the target fuel.

[0135] The beneficial effects of the above technical solution are: by using different sensors to collect and analyze the characteristic data of the target fuel in real time, the abnormal state detection and prediction of the target fuel can be realized, and the state alarm and early warning function of the target fuel can be realized through the generated fuel monitoring report, which can realize effective supervision of fuel quality and ensure the safety and high efficiency of fuel production.

[0136] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A fuel monitoring system, characterized in that: include: Data acquisition module: used for acquiring first characteristic data of target fuel in real time by using target sensor; Data processing module: used for preprocessing the first characteristic data to obtain target characteristic data; Monitoring module: used to detect and predict the abnormal state of the target fuel based on the analysis of the target characteristic data, and generate a corresponding fuel monitoring report; Alarm module: used to realize the status alarm and early warning function of the target fuel based on the fuel monitoring report; Wherein, the monitoring module includes: Annotation unit: used to compare the target characteristic data with the set characteristic threshold. If the target characteristic data exceeds the set characteristic threshold and the characteristic difference does not belong to the preset characteristic range, the current target characteristic data is determined to be abnormal, and the current target characteristic data is marked as a first-level abnormality; If the target characteristic data exceeds the set characteristic threshold and the characteristic difference falls within the preset characteristic range, the current target characteristic data is determined to be abnormal, and the current target characteristic data is marked as a secondary abnormality; If the target characteristic data does not exceed the set characteristic threshold, the current target characteristic data is marked as normal; Status Analysis Unit: Level judgment block: used to judge that the current target fuel state is normal if all target characteristic data of the target fuel at the current moment are normal; Otherwise, the abnormality assessment index is calculated based on the abnormality level of the target characteristic data at the current moment; The calculation formula of the abnormal assessment index is as follows: ; Wherein, y represents the abnormal evaluation index of the target fuel with abnormal target characteristic data; It is indicated as setting the first-level abnormal risk factor; It represents the number of target characteristic data with first-level anomalies; X represents the total number of target characteristic data; It is represented as the characteristic difference between the i-th target characteristic data marked with the first-level anomaly and the set characteristic threshold; It is represented as the influence weight of the i-th target characteristic data with the first-level anomaly marked on the fuel anomaly assessment; It is expressed as setting the secondary abnormal risk factor; It is expressed as the number of target characteristic data with marked secondary anomalies; It is represented as the characteristic difference between the i-th target characteristic data marked as the secondary anomaly and the set characteristic threshold; It is represented as the influence weight of the i-th target characteristic data of the marked secondary anomaly on the fuel anomaly assessment; Based on the set abnormal range table, according to the abnormal evaluation index, the abnormal level of the current target fuel is determined; Prediction block: used to predict the state of the target fuel at a preset future moment when the target fuel state is normal, and generate a state abnormality prediction result; Report generation unit: used to summarize the abnormal level, abnormal evaluation index, target characteristic data, characteristic difference and basic fuel information of the target fuel when the target fuel state is abnormal, and generate a fuel monitoring report; When the target fuel state is normal, the target characteristic data, characteristic difference, abnormal state prediction results and basic fuel information of the target fuel are summarized to generate a fuel monitoring report; Wherein, the prediction block includes: Model building sub-block: used to extract historical data collected by target sensors in different historical time periods from the equipment operation database to expand the data to obtain a training data set; Extracting the data time series features in the training data set to construct a first prediction training data feature set, and training a neural network to obtain a data prediction model; Extracting abnormal features from the training data set to construct a second prediction training data feature set, training a neural network, and obtaining an abnormal prediction model; Prediction sub-block: used to input the target characteristic data when the target fuel state is normal into the data prediction model, and output the possible characteristic value at a preset future moment; Anomaly detection sub-block: used to use the anomaly prediction model and the labeling unit to make two anomaly predictions on the target characteristic data state at the preset future moment based on the possible characteristic values ​​at the preset future moment, and analyze the two anomaly prediction results to obtain the state anomaly prediction result of the target fuel at the preset future moment; The anomaly detection sub-block includes: Inputting target characteristic data when the target fuel state is normal into the abnormality prediction model, and outputting a first abnormality prediction result of the target characteristic data at a preset future time; Using the labeling unit to perform anomaly detection on possible characteristic values ​​at a preset future time, and obtain a second anomaly prediction result corresponding to the target characteristic data; Compare the state determination result in the first abnormality prediction result with the state determination result in the second abnormality prediction result, and if the state determination results of the current target characteristic data are consistent, use the second abnormality prediction result as a reference abnormality prediction result of the current target characteristic data; If the state determination result of the current target characteristic data is inconsistent, the possible characteristic value of the current target characteristic data at the preset future time is adjusted to obtain a corrected characteristic value; Performing anomaly detection on the modified characteristic value by using a labeling unit to obtain a third anomaly prediction result corresponding to the target characteristic data, and using the third anomaly prediction result as a reference anomaly prediction result for the current target characteristic data; Using the state analysis unit to preset a reference abnormality prediction result at a future moment based on all target characteristic data of the target fuel, perform future state analysis on the fuel to obtain a first state prediction result; Determine the abnormal risk index of the target fuel at a preset future time based on the set risk factor of the abnormal level of the target characteristic data; Combining the first state prediction result with the abnormal risk index and outputting the result as the state abnormality prediction result; The calculation formula of the modified characteristic value is as follows: ; In the formula, It is expressed as the corrected characteristic value of the ath target characteristic data; It is represented as the possible characteristic value of the ath target characteristic data; It is represented as the characteristic value of the a-th target characteristic data at the current moment; Expressed as data prediction error compensation factor; It is expressed as the average of the absolute differences between the characteristic value of the a-th target characteristic data at the current moment and the characteristic values ​​at the previous three moments; It is expressed as the average of the absolute differences between the possible characteristic values ​​of the a-th target characteristic data and the characteristic values ​​at the current moment and the two moments before the current moment; It represents the data acquisition loss factor of the target sensor; Expressed as a data prediction bias compensation factor.

2. A fuel monitoring system according to claim 1, characterized in that: The first characteristic data includes sulfur content, ash content, and calorific value.

3. A fuel monitoring system according to claim 1, characterized in that: The alarm module comprises: Transmission unit: used for transmitting the corresponding fuel monitoring report to the alarm unit according to the target fuel state in the received fuel monitoring report, if the target fuel state is abnormal; If the target fuel status is normal, the corresponding fuel monitoring report is transmitted to the early warning unit; Alarm unit: used to extract the list of maintenance personnel for the target fuel from the fuel maintenance database according to the basic fuel information of the target fuel in the received fuel monitoring report; Filtering a target fuel status maintenance personnel who matches the abnormality level of the target fuel from the maintenance processing personnel list; Transmitting the fuel monitoring report to the target fuel status maintenance personnel for target fuel status maintenance; Early warning unit: used to generate a target fuel early warning task according to the received fuel monitoring report, if the target fuel state is abnormal at a preset future time, and add the target fuel early warning task to the early warning processing task column according to the abnormal risk index; The warning processing personnel maintain the fuel status according to the warning processing task list.

4. A fuel monitoring method, based on the fuel monitoring system according to any one of claims 1 to 3, wherein: include: Step 1: using a target sensor to collect first characteristic data of a target fuel in real time; Step 2: preprocessing the first characteristic data to obtain target characteristic data; Step 3: Based on the analysis of the target characteristic data, abnormal state detection and prediction of the target fuel are achieved, and a corresponding fuel monitoring report is generated; Step 4: Based on the fuel monitoring report, realize the status alarm and early warning function of the target fuel.

Images