CN113468798A - Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm - Google Patents

Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm Download PDF

Info

Publication number
CN113468798A
CN113468798A CN202110506800.0A CN202110506800A CN113468798A CN 113468798 A CN113468798 A CN 113468798A CN 202110506800 A CN202110506800 A CN 202110506800A CN 113468798 A CN113468798 A CN 113468798A
Authority
CN
China
Prior art keywords
parameter
parameters
coal mill
early warning
threshold interval
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110506800.0A
Other languages
Chinese (zh)
Inventor
王远鑫
许文良
陈俊
徐民
吴万范
程时鹤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Datang Boiler Pressure Vessel Examination Center Co Ltd
East China Electric Power Test Institute of China Datang Corp Science and Technology Research Institute Co Ltd
Original Assignee
Datang Boiler Pressure Vessel Examination Center Co Ltd
East China Electric Power Test Institute of China Datang Corp Science and Technology Research Institute Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Datang Boiler Pressure Vessel Examination Center Co Ltd, East China Electric Power Test Institute of China Datang Corp Science and Technology Research Institute Co Ltd filed Critical Datang Boiler Pressure Vessel Examination Center Co Ltd
Priority to CN202110506800.0A priority Critical patent/CN113468798A/en
Publication of CN113468798A publication Critical patent/CN113468798A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • Disintegrating Or Milling (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

The invention provides a medium speed coal mill fault early warning system based on a least square support vector machine algorithm, which comprises the following steps: (1) acquiring historical data of a coal mill, preprocessing the data, and selecting a parameter with larger correlation with a target regression parameter according to a Pearson correlation coefficient; (2) recording the selected historical parameters into an LS-SVM algorithm for training to obtain an optimal regression result model; (3) inputting the actual parameters into a trained algorithm to obtain an actual value regression result, calculating a self-adaptive threshold interval, judging whether the actual values are in the self-adaptive threshold interval, and early warning the abnormal state when the parameters are considered abnormal when the actual values continuously exceed the self-adaptive threshold interval for 15 s. The method can predict different faults according to different parameters.

Description

Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm
Technical Field
The invention relates to the technical field of equipment fault early warning, in particular to a medium speed coal mill fault early warning method and system based on a least square support vector machine algorithm.
Background
The coal mill of the thermal power plant is an important factor influencing the safe operation of a boiler as auxiliary equipment, along with the development of information technology, a DCS (distributed control system) of a power plant generates a large number of equipment operation parameters, and how to efficiently process and analyze the data resources is an important means for further improving the management level of a power station and guaranteeing the safe operation.
The abnormity or the fault of the equipment is shown by the change of the operation parameters in the operation process of the equipment, the characteristic parameters reflecting the equipment state are obtained by taking the main abnormity or the fault in the operation process of the equipment as clues, and the operation state of the equipment can be effectively evaluated.
Because the coal quality for combustion in a thermal power plant is complex and changeable, the medium-speed coal mill has the operation faults of abnormal vibration, blocked grinding, deflagration of a powder pipe, abnormal grinding body caused by foreign matters entering the coal mill during the operation, and the safe operation of the coal mill is seriously influenced. The thermal power plant mainly monitors the common operation safety risk by intermittent manual on-site inspection and centralized control room dial plate monitoring, and the intermittent manual on-site inspection and the centralized control room dial plate monitoring often cannot find and early warn the fault risk in time. The medium-speed coal mill is a direct-blowing coal mill, the coal powder stays in the coal mill for a short time without being stored, and the parameters of the coal mill can reflect the real-time state. This feature is not currently well utilized in the early warning of faults in medium speed coal mills.
Such as the vibration fault diagnosis method of the medium speed coal mill disclosed in the application number CN201510170470.7, the method comprises the steps of firstly analyzing the fault characteristics of the vibration of the coal mill to obtain the current of the coal mill as an important monitoring parameter for analyzing the vibration fault of the coal mill, then screening out four auxiliary variables of the air pressure at a mill outlet, the coal feeding quantity of the coal feeder, the primary air quantity at the mill inlet and the temperature at the mill outlet from the analysis through the correlation of historical data, predicting the current by the four auxiliary variables, taking the difference between the actually measured current and the predicted current, taking the current residual sequence of the coal mill to construct the grinding vibration quantity, and performing three-layer wavelet packet decomposition on the grinding current residual sequence to obtain an energy ratio of 8 frequency bands, and performing statistical analysis on the energy ratio of two faults to obtain the characteristic quantity of serious grinding roller abrasion and obvious characteristic quantity distinguishing of two faults of foreign matters entering the grinding roller. According to the method, only the current of the coal mill is predicted, the vibration fault of the coal mill is judged according to the prediction result, and other faults cannot be predicted by using other parameters.
Disclosure of Invention
The invention aims to solve the technical problem of predicting various faults through the real-time parameter relation of the coal mill.
The invention solves the technical problems through the following technical means:
the medium speed coal mill fault early warning method based on the least square support vector machine algorithm comprises the following steps:
step 1, collecting historical data of a coal mill, carrying out standardization processing on the data, and selecting parameters with large correlation with target parameters according to a Pearson correlation system;
step 2, inputting the selected historical parameters into an LS-SVM algorithm for training to obtain an optimal regression result model;
step 3, inputting each actual parameter into a trained algorithm to obtain an actual value regression result, and calculating a self-adaptive threshold interval;
and 4, judging whether the actual value is in the adaptive threshold interval, continuously setting the time length to exceed the adaptive threshold interval, considering that the parameter is abnormal, and early warning the abnormal state.
Aiming at the medium-speed coal mill, parameters with high correlation with target parameters are selected through Pearson correlation coefficients, the regression target parameters are regressed by using an LS-SVM algorithm, the regression target parameters are compared with the current target parameters, threshold upper and lower limits are set for the regression parameters, and real-time target parameters and threshold upper and lower limit intervals are compared, so that threshold interval distortion caused by the fact that the running state of the auxiliary machine deviates from a normal state to different degrees in different variable working condition stages is avoided.
Further, the step 1 specifically comprises: firstly, standardizing all parameters of collected historical data, wherein the standardized formula is as follows:
Figure BDA0003058718800000021
where x (t) denotes the actual value at the moment of the respective parameter t,
Figure BDA0003058718800000022
is the mean value of each parameter, σx(t)The variance of each parameter;
calculating the Pearson correlation coefficient between the target parameter of the coal mill and each parameter, wherein the calculation formula is as follows:
Figure BDA0003058718800000023
wherein X is other parameter value, Y is target parameter value, Cov (X, Y) is covariance of X and Y, D (X) is variance of X, and D (Y) is variance of Y.
Further, the step 2 specifically comprises: with ytIs a target parameter value, xtInputting LS-SVM for other parameters for training:
Figure BDA0003058718800000024
Figure BDA0003058718800000025
the Lagrange multiplier method is adopted to convert the original problem into a single parameter, and the new problem is as follows:
Figure BDA0003058718800000031
for omega, b, e respectivelyt，atThe derivative is equal to 0:
Figure BDA0003058718800000032
Figure BDA0003058718800000033
Figure BDA0003058718800000034
Figure BDA0003058718800000035
according to the above conditions, a can be listedtAnd b system of linear equations
yt T·at＝[0,0,0...,0]The number of the N is N in total,
Figure BDA0003058718800000036
the number of the N is N in total,
solving the above equation can obtain at，b
Wherein gamma is a weight value and is used for finding the optimal hyperplane in a balanced way and minimizing the deviation amount;
Figure BDA0003058718800000037
is xtSelf inner product; y isk、yl，xk、xlAnd representing target parameter values and other parameter values of the historical data at the time k and the time l respectively, and adjusting gamma according to the regression result.
Further, the step 3 specifically includes:
step 31. according to the current input parameters, the formula
Figure BDA0003058718800000038
Wherein K (x)t,xAt present) The inner product of the historical input parameter and the current input parameter;
step 32, calculating the average value epsilon of the previous K predicted valuesKAnd variance deltaKWith an adaptive threshold interval of [ epsilon ]K-zδK,εK+zδK]Wherein Z is a selected integer.
Corresponding to the method, the invention also provides a medium speed coal mill fault early warning system based on the least square support vector machine algorithm, which comprises the following steps:
the preprocessing module is used for acquiring historical data of the coal mill, carrying out standardization processing on the data, and selecting parameters with large correlation with target parameters according to a Pearson correlation system;
the model training module is used for inputting the selected historical parameters into an LS-SVM algorithm for training to obtain an optimal regression result model;
the self-adaptive threshold interval calculation module is used for inputting each actual parameter into a trained algorithm to obtain an actual value regression result and calculating a self-adaptive threshold interval;
and the judging module is used for judging whether the actual value is in the self-adaptive threshold interval or not, continuously setting the time length to exceed the self-adaptive threshold interval, considering that the parameter is abnormal, and early warning the abnormal state.
Further, the execution process of the preprocessing module specifically includes: firstly, standardizing all parameters of collected historical data, wherein the standardized formula is as follows:
Figure BDA0003058718800000041
where x (t) denotes the actual value at the moment of the respective parameter t,
Figure BDA0003058718800000042
is the mean value of each parameter, σx(t)For each parameterA difference;
calculating the Pearson correlation coefficient between the target parameter of the coal mill and each parameter, wherein the calculation formula is as follows:
Figure BDA0003058718800000043
wherein X is other parameter value, Y is target parameter value, Cov (X, Y) is covariance of X and Y, D (X) is variance of X, and D (Y) is variance of Y.
Further, the model training module specifically executes a process that: with ytIs a target parameter value, xtInputting LS-SVM for other parameters for training:
Figure BDA0003058718800000044
Figure BDA0003058718800000045
the Lagrange multiplier method is adopted to convert the original problem into a single parameter, and the new problem is as follows:
Figure BDA0003058718800000046
for omega, b, e respectivelyt，atThe derivative is equal to 0:
Figure BDA0003058718800000051
Figure BDA0003058718800000052
Figure BDA0003058718800000053
Figure BDA0003058718800000054
according to the above conditions, a can be listedtAnd b system of linear equations
yt T·at＝[0,0,0...,0]The number of the N is N in total,
Figure BDA0003058718800000055
the number of the N is N in total,
solving the above equation can obtain at，b
Wherein gamma is a weight value and is used for finding the optimal hyperplane in a balanced way and minimizing the deviation amount;
Figure BDA0003058718800000056
is xtSelf inner product; y isk、yl，xk、xlAnd representing target parameter values and other parameter values of the historical data at the time k and the time l respectively, and adjusting gamma according to the regression result.
Further, the adaptive threshold interval calculation module specifically executes a process including:
step 31. according to the current input parameters, the formula
Figure BDA0003058718800000057
Wherein K (x)t,xAt present) The inner product of the historical input parameter and the current input parameter;
step 32, calculating the average value epsilon of the previous K predicted valuesKAnd variance deltaKWith an adaptive threshold interval of [ epsilon ]K-zδK,εK+zδK]Wherein Z is a selected integer.
The present invention also provides a processing device comprising at least one processor, and at least one memory communicatively coupled to the processor, wherein: the memory stores program instructions executable by the processor, which when called by the processor are capable of performing the methods described above.
The present invention also provides a computer-readable storage medium storing computer instructions that cause the computer to perform the above-described method.
The invention has the advantages that:
aiming at the medium-speed coal mill, parameters with high correlation with target parameters are selected through Pearson correlation coefficients, the regression target parameters are regressed by using an LS-SVM algorithm, the regression target parameters are compared with the current target parameters, threshold upper and lower limits are set for the regression parameters, and real-time target parameters and threshold upper and lower limit intervals are compared, so that threshold interval distortion caused by the fact that the running state of the auxiliary machine deviates from a normal state to different degrees in different variable working condition stages is avoided.
The invention carries out regression prediction on main parameters of the medium-speed coal mill, such as current, inlet air quantity, outlet air pressure, outlet temperature, pulverized coal pipe air speed, mill body differential pressure and the like, and carries out early warning diagnosis on various faults (coal breaking, blockage, pulverized coal leakage of the pulverized coal pipe and the like) of the coal mill except vibration faults.
Drawings
Fig. 1 is a flow chart of a medium speed coal mill fault early warning algorithm based on a least square support vector machine algorithm in the embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, the medium speed coal mill fault early warning method based on the least square support vector machine algorithm includes the following steps:
step 1, collecting historical data of a coal mill, carrying out standardization processing on the data, and selecting parameters with large correlation with target parameters according to a Pearson correlation system; the method specifically comprises the following steps:
taking the current of the coal mill as an example, acquiring historical data of the coal mill, carrying out standardization processing on the acquired data, and calculating a parameter with larger correlation with the current of the coal mill according to a Pearson correlation coefficient;
the selected coal mill parameters include: the current (A), the coal feeding amount (t/h), the inlet air pressure (KPa), the differential pressure (KPa) of a coal mill, the inlet air temperature (DEG C), the outlet air temperature (DEG C), the inlet air amount (t/h), the load (MW), the air speed (m/s) of a powder pipe 1, the air speed (m/s) of a powder pipe 2, the air speed (m/s) of a powder pipe 3, the air speed (m/s) of a powder pipe 4, the vibration (mm) of a bearing x direction, the grinding roller temperature (DEG C) 2, the grinding roller temperature (DEG C), the CO concentration (ppm) of a grinding body, the sealing air pressure (KPa), the primary air differential pressure (KPa) of sealing air and the loading oil pressure feedback of the grinding roller are totally 20 items.
The method comprises the following steps of standardizing all parameters of collected historical data, wherein a standardization formula is as follows: firstly, standardizing all parameters of collected historical data, wherein the standardized formula is as follows:
Figure BDA0003058718800000071
where x (t) denotes the actual value at the moment of the respective parameter t,
Figure BDA0003058718800000072
is the mean value of each parameter, σx(t)The variance of each parameter;
calculating the Pearson correlation coefficient between the target parameter of the coal mill and each parameter, wherein the calculation formula is as follows:
Figure BDA0003058718800000073
wherein X is other parameter value, Y is target parameter value, Cov (X, Y) is covariance of X and Y, D (X) is variance of X, and D (Y) is variance of Y.
The parameters having a large current dependency include: coal quantity, inlet wind pressure, outlet wind pressure, inlet temperature, load, sealing wind pressure and grinding roller loading oil pressure.
Step 2, inputting the selected historical parameters into an LS-SVM algorithm for training to obtain an optimal regression result model; the method specifically comprises the following steps: using current as target value, coal quantity, inlet wind pressure, outlet wind pressure, inlet temperature, load, sealing wind pressure and grinding roller loading oil pressure as input parameters, and using ytIs a target parameter value, xtInputting LS-SVM for other parameters for training:
Figure BDA0003058718800000074
Figure BDA0003058718800000075
the Lagrange multiplier method is adopted to convert the original problem into a single parameter, and the new problem is as follows:
Figure BDA0003058718800000076
for omega, b, e respectivelyt，atThe derivative is equal to 0:
Figure BDA0003058718800000077
Figure BDA0003058718800000078
Figure BDA0003058718800000079
Figure BDA00030587188000000710
according to the above conditions, a can be listedtAnd b system of linear equations
yt T·at＝[0,0,0...,0]The number of the N is N in total,
Figure BDA0003058718800000081
the number of the N is N in total,
solving the above equation can obtain at，b
Wherein gamma is a weight value and is used for finding the optimal hyperplane in a balanced way and minimizing the deviation amount;
Figure BDA0003058718800000082
is xtSelf inner product; y isk、yl，xk、xlAnd representing target parameter values and other parameter values of the historical data at the time k and the time l respectively, and adjusting gamma according to the regression result.
Step 3, inputting each actual parameter into a trained algorithm to obtain an actual value regression result, and calculating a self-adaptive threshold interval; the method specifically comprises the following steps: inputting an actual input value to the trained LS-SVM model, regressing an actual current value, and calculating an adaptive threshold interval according to a regression result:
step 31. according to the current input parameters, the formula
Figure BDA0003058718800000083
Wherein K (x)t,xAt present) The inner product of the historical input parameter and the current input parameter;
step 32, calculating the average value epsilon of the previous K predicted valuesKAnd variance deltaKWith an adaptive threshold interval of [ epsilon ]K-zδK,εK+zδK]Wherein Z is a selected integer.
And 4, judging whether the actual value is in the adaptive threshold interval, continuously setting the time length to exceed the adaptive threshold interval, considering that the parameter is abnormal, and early warning the abnormal state. In the embodiment, the condition that the parameter is considered to be abnormal when the continuous 15s exceeds the adaptive threshold interval is provided, and the abnormal state is warned.
Aiming at the medium-speed coal mill, parameters with high correlation with target parameters are selected through Pearson correlation coefficients, the regression target parameters are regressed by using an LS-SVM algorithm, the regression target parameters are compared with the current target parameters, threshold upper and lower limits are set for the regression parameters, and real-time target parameters and threshold upper and lower limit intervals are compared, so that threshold interval distortion caused by the fact that the running state of the auxiliary machine deviates from a normal state to different degrees in different variable working condition stages is avoided.
The invention carries out regression prediction on main parameters of the medium-speed coal mill, such as current, inlet air quantity, outlet air pressure, outlet temperature, pulverized coal pipe air speed, mill body differential pressure and the like, and carries out early warning diagnosis on various faults (coal breaking, blockage, pulverized coal leakage of the pulverized coal pipe and the like) of the coal mill except vibration faults.
Corresponding to the method, the invention also provides a medium speed coal mill fault early warning system based on the least square support vector machine algorithm, which is characterized by comprising the following steps:
the preprocessing module is used for acquiring historical data of the coal mill, carrying out standardization processing on the data, and selecting parameters with large correlation with target parameters according to a Pearson correlation system; taking the current of the coal mill as an example, acquiring historical data of the coal mill, carrying out normalization processing on the acquired data, and calculating a parameter with larger correlation with the current of the coal mill according to a Pearson correlation coefficient;
the selected coal mill parameters include: the current (A), the coal feeding amount (t/h), the inlet air pressure (KPa), the differential pressure (KPa) of a coal mill, the inlet air temperature (DEG C), the outlet air temperature (DEG C), the inlet air amount (t/h), the load (MW), the air speed (m/s) of a powder pipe 1, the air speed (m/s) of a powder pipe 2, the air speed (m/s) of a powder pipe 3, the air speed (m/s) of a powder pipe 4, the vibration (mm) of a bearing x direction, the grinding roller temperature (DEG C) 2, the grinding roller temperature (DEG C), the CO concentration (ppm) of a grinding body, the sealing air pressure (KPa), the primary air differential pressure (KPa) of sealing air and the loading oil pressure feedback of the grinding roller are totally 20 items.
The method comprises the following steps of standardizing all parameters of collected historical data, wherein a standardization formula is as follows: firstly, standardizing all parameters of collected historical data, wherein the standardized formula is as follows:
Figure BDA0003058718800000091
where x (t) denotes the actual value at the moment of the respective parameter t,
Figure BDA0003058718800000092
is the mean value of each parameter, σx(t)The variance of each parameter;
calculating the Pearson correlation coefficient between the target parameter of the coal mill and each parameter, wherein the calculation formula is as follows:
Figure BDA0003058718800000093
wherein X is other parameter value, Y is target parameter value, Cov (X, Y) is covariance of X and Y, D (X) is variance of X, and D (Y) is variance of Y.
The parameters having a large current dependency include: coal quantity, inlet wind pressure, outlet wind pressure, inlet temperature, load, sealing wind pressure and grinding roller loading oil pressure.
The model training module is used for inputting the selected historical parameters into an LS-SVM algorithm for training to obtain an optimal regression result model; the specific execution process comprises the following steps:
using current as target value, coal quantity, inlet wind pressure, outlet wind pressure, inlet temperature, load, sealing wind pressure and grinding roller loading oil pressure as input parameters, and using ytIs a target parameter value, xtInputting LS-SVM for other parameters for training:
Figure BDA0003058718800000094
Figure BDA0003058718800000095
the Lagrange multiplier method is adopted to convert the original problem into a single parameter, and the new problem is as follows:
Figure BDA0003058718800000101
for omega, b, e respectivelyt，atThe derivative is equal to 0:
Figure BDA0003058718800000102
Figure BDA0003058718800000103
Figure BDA0003058718800000104
Figure BDA0003058718800000105
according to the above conditions, a can be listedtAnd b system of linear equations
yt T·at＝[0,0,0...,0]The number of the N is N in total,
Figure BDA0003058718800000106
the number of the N is N in total,
solving the above equation can obtain at，b
Wherein gamma is a weight value and is used for finding the optimal hyperplane in a balanced way and minimizing the deviation amount;
Figure BDA0003058718800000107
is xtSelf inner product; y isk、yl，xk、xlAnd representing target parameter values and other parameter values of the historical data at the time k and the time l respectively, and adjusting gamma according to the regression result.
The self-adaptive threshold interval calculation module is used for inputting each actual parameter into a trained algorithm to obtain an actual value regression result and calculating a self-adaptive threshold interval; the method specifically comprises the following steps: inputting an actual input value to the trained LS-SVM model, regressing an actual current value, and calculating an adaptive threshold interval according to a regression result:
step 31. according to the current input parameters, the formula
Figure BDA0003058718800000108
Wherein K (x)t,xAt present) The inner product of the historical input parameter and the current input parameter;
step 32, calculating the average value epsilon of the previous K predicted valuesKAnd variance deltaKWith an adaptive threshold interval of [ epsilon ]K-zδK,εK+zδK]Wherein Z is a selected integer.
And the judging module is used for judging whether the actual value is in the adaptive threshold interval or not, continuously setting the time length to exceed the adaptive threshold interval, considering that the parameter is abnormal, early warning the abnormal state, and specifically executing the step 4.
The present invention also provides a processing device comprising at least one processor, and at least one memory communicatively coupled to the processor, wherein: the memory stores program instructions executable by the processor, which when called by the processor are capable of performing the methods described above.
The present invention also provides a computer-readable storage medium storing computer instructions that cause the computer to perform the above-described method.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. The medium speed coal mill fault early warning method based on the least square support vector machine algorithm is characterized by comprising the following steps of:
step 1, collecting historical data of a coal mill, carrying out standardization processing on the data, and selecting parameters with large correlation with target parameters according to a Pearson correlation system;
step 2, inputting the selected historical parameters into an LS-SVM algorithm for training to obtain an optimal regression result model;
step 3, inputting each actual parameter into a trained algorithm to obtain an actual value regression result, and calculating a self-adaptive threshold interval;
and 4, judging whether the actual value is in the adaptive threshold interval, continuously setting the time length to exceed the adaptive threshold interval, considering that the parameter is abnormal, and early warning the abnormal state.
2. The medium speed coal mill fault early warning method based on the least square support vector machine algorithm according to claim 1, wherein the step 1 specifically comprises: firstly, standardizing all parameters of collected historical data, wherein the standardized formula is as follows:
Figure FDA0003058718790000011
where x (t) denotes the actual value at the moment of the respective parameter t,
Figure FDA0003058718790000012
is the mean value of each parameter, σx(t)The variance of each parameter;
calculating the Pearson correlation coefficient between the target parameter of the coal mill and each parameter, wherein the calculation formula is as follows:
Figure FDA0003058718790000013
wherein X is other parameter value, Y is target parameter value, Cov (X, Y) is covariance of X and Y, D (X) is variance of X, and D (Y) is variance of Y.
3. The method of claim 2 based on a least squares support vector machine algorithmThe coal mill fault early warning method is characterized in that the step 2 specifically comprises the following steps: with ytIs a target parameter value, xtInputting LS-SVM for other parameters for training:
Figure FDA0003058718790000014
Figure FDA0003058718790000015
the Lagrange multiplier method is adopted to convert the original problem into a single parameter, and the new problem is as follows:
Figure FDA0003058718790000016
for omega, b, e respectivelyt，atThe derivative is equal to 0:
Figure FDA0003058718790000021
Figure FDA0003058718790000022
Figure FDA0003058718790000023
Figure FDA0003058718790000024
according to the above conditions, a can be listedtAnd b system of linear equations
yt T·at＝[0,0,0...,0]The number of the N is N in total,
Figure FDA0003058718790000025
the number of the N is N in total,
solving the above equation can obtain at，b
Wherein gamma is a weight value and is used for finding the optimal hyperplane in a balanced way and minimizing the deviation amount;
Figure FDA0003058718790000026
is xtSelf inner product; y isk、yl，xk、xlAnd representing target parameter values and other parameter values of the historical data at the time k and the time l respectively, and adjusting gamma according to the regression result.
4. The medium speed coal mill fault early warning method based on the least square support vector machine algorithm according to claim 3, wherein the step 3 specifically comprises:
step 31. according to the current input parameters, the formula
Figure FDA0003058718790000027
Wherein K (x)t,xAt present) The inner product of the historical input parameter and the current input parameter;
step 32, calculating the average value epsilon of the previous K predicted valuesKAnd variance deltaKWith an adaptive threshold interval of [ epsilon ]K-zδK,εK+zδK]Wherein Z is a selected integer.
5. Medium speed pulverizer fault early warning system based on least square support vector machine algorithm, its characterized in that includes:
the preprocessing module is used for acquiring historical data of the coal mill, normalizing the data and selecting parameters with large correlation with target parameters according to a Pearson correlation system;
the model training module is used for inputting the selected historical parameters into an LS-SVM algorithm for training to obtain an optimal regression result model;
the self-adaptive threshold interval calculation module is used for inputting each actual parameter into a trained algorithm to obtain an actual value regression result and calculating a self-adaptive threshold interval;
and the judging module is used for judging whether the actual value is in the self-adaptive threshold interval or not, continuously setting the time length to exceed the self-adaptive threshold interval, considering that the parameter is abnormal, and early warning the abnormal state.
6. The medium speed coal mill fault early warning system based on the least squares support vector machine algorithm according to claim 5, wherein the preprocessing module specifically executes the following process: firstly, standardizing all parameters of collected historical data, wherein the standardized formula is as follows:
Figure FDA0003058718790000031
where x (t) denotes the actual value at the moment of the respective parameter t,
Figure FDA0003058718790000032
is the mean value of each parameter, σx(t)The variance of each parameter;
calculating the Pearson correlation coefficient between the target parameter of the coal mill and each parameter, wherein the calculation formula is as follows:
Figure FDA0003058718790000033
wherein X is other parameter value, Y is target parameter value, Cov (X, Y) is covariance of X and Y, D (X) is variance of X, and D (Y) is variance of Y.
7. The coal mill fault early warning method based on the least squares support vector machine algorithm according to claim 6, wherein the model training module is specifically executed in the following process: with ytIs a target parameter value, xtInputting LS-SVM for other parameters for training:
Figure FDA0003058718790000034
Figure FDA0003058718790000035
the Lagrange multiplier method is adopted to convert the original problem into a single parameter, and the new problem is as follows:
Figure FDA0003058718790000036
for omega, b, e respectivelyt，atThe derivative is equal to 0:
Figure FDA0003058718790000041
Figure FDA0003058718790000042
Figure FDA0003058718790000043
Figure FDA0003058718790000044
according to the above conditions, a can be listedtAnd b system of linear equations
yt T·at＝[0,0,0...,0]The number of the N is N in total,
Figure FDA0003058718790000045
the number of the N is N in total,
solving the above equation can obtain at，b
Wherein gamma is a weight value and is used for finding the optimal hyperplane in a balanced way and minimizing the deviation amount;
Figure FDA0003058718790000046
is xtSelf inner product; y isk、yl，xk、xlAnd representing target parameter values and other parameter values of the historical data at the time k and the time l respectively, and adjusting gamma according to the regression result.
8. The medium speed coal mill fault early warning system based on least squares support vector machine algorithm according to claim 7, characterized in that the adaptive threshold interval calculation module specifically executes the process:
step 31. according to the current input parameters, the formula
Figure FDA0003058718790000047
Wherein K (x)t,xAt present) The inner product of the historical input parameter and the current input parameter;
step 32, calculating the average value epsilon of the previous K predicted valuesKAnd variance deltaKWith an adaptive threshold interval of [ epsilon ]K-zδK,εK+zδK]Wherein Z is a selected integer.
9. A processing device comprising at least one processor and at least one memory communicatively coupled to the processor, wherein: the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1 to 4.
10. A computer-readable storage medium storing computer instructions for causing a computer to perform the method of any one of claims 1 to 4.
CN202110506800.0A 2021-05-10 2021-05-10 Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm Pending CN113468798A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110506800.0A CN113468798A (en) 2021-05-10 2021-05-10 Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110506800.0A CN113468798A (en) 2021-05-10 2021-05-10 Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm

Publications (1)

Publication Number Publication Date
CN113468798A true CN113468798A (en) 2021-10-01

Family

ID=77870608

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110506800.0A Pending CN113468798A (en) 2021-05-10 2021-05-10 Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm

Country Status (1)

Country Link
CN (1) CN113468798A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114492938A (en) * 2021-12-29 2022-05-13 中国大唐集团科学技术研究总院有限公司华东电力试验研究院 Coal mill fault early warning method and system based on BPNN model and adaptive threshold
CN115097301A (en) * 2022-06-22 2022-09-23 佛山众陶联供应链服务有限公司 Ceramic raw material workshop intermittent ball mill early warning method and system
CN117111568A (en) * 2023-10-24 2023-11-24 安联奇智(安徽)科技有限公司 Equipment monitoring method, device, equipment and storage medium based on Internet of things

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114492938A (en) * 2021-12-29 2022-05-13 中国大唐集团科学技术研究总院有限公司华东电力试验研究院 Coal mill fault early warning method and system based on BPNN model and adaptive threshold
CN115097301A (en) * 2022-06-22 2022-09-23 佛山众陶联供应链服务有限公司 Ceramic raw material workshop intermittent ball mill early warning method and system
CN117111568A (en) * 2023-10-24 2023-11-24 安联奇智(安徽)科技有限公司 Equipment monitoring method, device, equipment and storage medium based on Internet of things
CN117111568B (en) * 2023-10-24 2024-01-19 安联奇智(安徽)科技有限公司 Equipment monitoring method, device, equipment and storage medium based on Internet of things

Similar Documents

Publication Publication Date Title
CN113468798A (en) Medium-speed coal mill fault early warning method and system based on least square support vector machine algorithm
CN110298455B (en) Mechanical equipment fault intelligent early warning method based on multivariate estimation prediction
CN110346160B (en) SVR-based coal mill efficiency abnormity detection and diagnosis method
CN111308991B (en) Coal mill operation fault identification method and application
CN111275288B (en) XGBoost-based multidimensional data anomaly detection method and device
CN109240244B (en) Data-driven equipment running state health degree analysis method and system
CN108681633B (en) Condensate pump fault early warning method based on state parameters
CN113705924B (en) Intelligent diagnosis method and system for thermal control equipment
US8606544B2 (en) Methods and systems for detecting deviation of a process variable from expected values
US20080183427A1 (en) Heat Exchanger Fouling Detection
EP3296822B1 (en) Model-plant mismatch detection using model parameter data clustering for paper machines or other systems
CN114757380B (en) Fault early warning system and method for thermal power plant, electronic equipment and storage medium
CN110209144B (en) Two-layer real-time monitoring and alarm tracing method based on dynamic and static cooperative difference analysis
CN114595623A (en) XGboost algorithm-based unit equipment reference value prediction method and system
CN109345143B (en) Intelligent fan running state evaluation method and device and wind turbine generator
Abbasi et al. Development of predictive maintenance interface using multiple linear regression
CN113505497B (en) Method and system for monitoring slurry quality of wet flue gas desulfurization absorption tower
Yuan et al. Analysis of multivariable control performance assessment techniques
Wang Key techniques in intelligent predictive maintenance (IPdM)–a framework of intelligent faults diagnosis and prognosis system (IFDaPS)
Chesterman et al. Condition monitoring of wind turbines and extraction of healthy training data using an ensemble of advanced statistical anomaly detection models
CN109784127B (en) Equipment health state early warning method and system thereof
CN116756704A (en) Method for predicting coal quantity into furnace by using wind powder measurement parameters
CN115456041A (en) Equipment fault early warning method and device, computing equipment and storage medium
EP4045791B1 (en) Method and an apparatus for computer-implemented monitoring of a wind turbine
CN111679648B (en) Multivariate closed-loop control loop performance evaluation method based on Gaussian process regression

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