CN117093836A - Last-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM - Google Patents
Last-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM Download PDFInfo
- Publication number
- CN117093836A CN117093836A CN202310847975.7A CN202310847975A CN117093836A CN 117093836 A CN117093836 A CN 117093836A CN 202310847975 A CN202310847975 A CN 202310847975A CN 117093836 A CN117093836 A CN 117093836A
- Authority
- CN
- China
- Prior art keywords
- steam temperature
- vmd
- early warning
- lstm
- final
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 241000283153 Cetacea Species 0.000 claims abstract description 30
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 30
- 238000005259 measurement Methods 0.000 claims abstract description 11
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 16
- 238000005457 optimization Methods 0.000 claims description 15
- 239000003245 coal Substances 0.000 claims description 11
- 239000000779 smoke Substances 0.000 claims description 8
- 238000003062 neural network model Methods 0.000 claims description 7
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 6
- 238000005507 spraying Methods 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 239000000446 fuel Substances 0.000 claims description 5
- 238000003860 storage Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 238000007599 discharging Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 description 4
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000003247 decreasing effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 239000003546 flue gas Substances 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
- G06F18/15—Statistical pre-processing, e.g. techniques for normalisation or restoring missing data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Control Of Turbines (AREA)
Abstract
The application discloses a final-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM, comprising the following steps: acquiring measuring point history data of the steam temperature of the outlet of the final-stage reheater in a normal working state, and performing VMD variation modal decomposition on the measuring point history data; reconstructing and denoising the measuring point signals according to VMD variation modal decomposition results and combining the original signal correlation degree of the measuring points to obtain a denoised historical data set; preprocessing a historical data set, establishing a reheater outlet steam temperature prediction model based on preprocessed measurement point data, and optimizing super-parameters of the model by improving whale algorithm; and calculating a residual sequence of a model prediction result and a reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and alarming the final-stage reheater air temperature when the average value of the final-stage reheater outlet steam temperature residual is larger than the set early warning threshold. The application can early warn in time before the outlet steam temperature of the final-stage reheater exceeds the allowable range, thereby controlling the reheat steam temperature in advance.
Description
Technical Field
The application relates to the technical field of thermal power plant final-stage reheater fault early warning, in particular to a final-stage reheater steam temperature early warning method and system based on a variation modal decomposition combined improved whale algorithm optimization long-short-term memory neural network (VMD-EWOA-LSTM).
Background
In a thermal power generation unit, a boiler is an important power plant capable of converting thermal energy of fuel into kinetic energy of steam. The final reheater is used as an indispensable part of a steam-water system of a boiler, and can heat the exhaust steam of a high-pressure cylinder of a steam turbine into reheated steam and then send the reheated steam into a middle-pressure cylinder and a low-pressure cylinder of the steam turbine to continue expansion and work, and the process can improve the cycle efficiency by 4% -5%. However, when factors such as boiler load, feedwater temperature, coal type, air quantity and the like are changed, the outlet steam temperature of the final stage reheater of the boiler is extremely liable to fluctuate. When the temperature of steam at the outlet of the final-stage reheater is too high, the service life of a reheater pipeline can be shortened, and the maximum heat resistance strength of a steam turbine cylinder, a nozzle and a blade can be exceeded; when the temperature of steam at the outlet of the final-stage reheater is too low, the efficiency of the steam turbine can be reduced, the coal consumption can be increased, the blades of the steam turbine can be corroded, and the safe operation of the steam turbine is endangered.
At present, in a thermal power generating unit, the outlet steam temperature of a reheater is controlled by methods of adjusting the opening of a flue gas side baffle, changing the inclination angle of a burner, recycling flue gas, reducing the temperature of accident water spraying, and the like, but the control methods such as the methods of adjusting the opening of the flue gas side baffle, changing the inclination angle of the burner, reducing the temperature of the water spraying, and the like have certain hysteresis. The existing deep learning-based method such as a long-short-term memory network LSTM is widely applied to the field of prediction of reheat steam temperature of a thermal power plant, however, a measuring point signal acquired in the thermal power plant is inevitably doped with environmental noise and other interference signals, and a reheat steam temperature prediction model established by using an original signal of the measuring point is high in randomness and low in robustness. And the prediction effect and generalization capability of the LSTM neural network are greatly dependent on the setting of super parameters, and the prediction effect is unstable due to the fact that parameters are manually set.
Disclosure of Invention
This section is intended to outline some aspects of embodiments of the application and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section as well as in the description of the application and in the title of the application, which may not be used to limit the scope of the application.
The present application has been made in view of the above-described problems occurring in the prior art.
Therefore, the application provides a final-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM, which solve the problems of strong randomness, low robustness and unstable prediction effect of the existing reheat steam temperature prediction.
In order to solve the technical problems, the application provides the following technical scheme:
in a first aspect, an embodiment of the present application provides a final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, including:
acquiring measuring point historical data of steam temperature of an outlet of a final-stage reheater in a normal working state, and performing VMD variation modal decomposition on the measuring point historical data;
reconstructing and denoising the measuring point signals according to the VMD variation modal decomposition result and combining the original signal correlation degree of the measuring point to obtain a denoised historical data set;
preprocessing the historical data set, establishing a reheater outlet steam temperature prediction model based on preprocessed measuring point data, and optimizing super parameters of the model by improving whale algorithm;
and calculating a residual sequence of the model prediction result and the reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and alarming the temperature of the final-stage reheater when the average value of the final-stage reheater outlet steam temperature residual is larger than the set early warning threshold.
As a preferable scheme of the final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, the application comprises the following steps: the measuring point history data at least comprises: coal amount, smoke discharging temperature, total fuel amount, total air quantity, main steam pressure, opening of a secondary water spraying valve, opening of a smoke baffle plate and outlet pressure of a reheater.
As a preferable scheme of the final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, the application comprises the following steps: performing VMD variation modal decomposition on the measurement point data comprises:
establishing a variation constraint model according to the acquired measurement point historical data, and introducing a penalty factor and a Lagrange operator to convert the constraint model into a non-constraint model;
and solving saddle point solutions of the unconstrained model by using an overlap direction multiplier method to obtain each IMF component.
As a preferable scheme of the final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, the application comprises the following steps: and solving saddle point solutions of the unconstrained model to obtain IMF components, wherein the method comprises the following steps of:
will be{ω k 1 }、{λ 1 Initializing the number K of modes and the bandwidth alpha to 0;
for a pair ofω k 1 And lambda (lambda) 1 Iterative updating is performed, expressed as:
where k is the number of decomposition layers, α is the bandwidth, λ is the Lagrangian operator, { u k The k-th IMF mode function, { ω k The center frequency corresponding to the kth IMF mode function is represented, f represents the original signal, and tau is the time scale of the mode function;
the iteration termination condition is expressed as:
wherein epsilon is the judgment precision;
when the iteration result is smaller than the judgment precision, outputting K modal components;
and when the iteration result is not less than the judgment precision, continuing to update the iteration.
As a preferable scheme of the final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, the application comprises the following steps: the reconstruction denoising of the measuring point signal by combining the original signal correlation degree of the measuring point comprises the following steps:
when the Spekerman coefficient of the IMF component and the original signal of the measuring point is smaller than 0.1, eliminating the IMF noise component;
when the Spekerman coefficient of the IMF component and the original signal of the measuring point is not less than 0.1, reserving the IMF component;
and adding the reserved IMF components to obtain a noise-reduced historical data set.
As a preferable scheme of the final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, the application comprises the following steps: establishing a reheater outlet steam temperature prediction model based on the preprocessed measuring point data, and optimizing super parameters of the model by improving whale algorithm, wherein the method comprises the following steps: taking the normalized measuring point data as an input variable of a reheater outlet steam temperature prediction model;
optimizing LSTM neural network model parameters by using whale optimization algorithm with self-adaptive weight and variation factor, wherein the parameters of the optimized LSTM neural network model at least comprise an initialization learning rate, the number of hidden layers and a regularization coefficient;
the fitness function of the whale optimizing algorithm is the root mean square error of the reheat steam temperature predicted value and the actual value;
outputting a reheat steam temperature predicted value when the training result of the reheater outlet steam temperature predicted model meets a preset target; otherwise, continuing to optimize the LSTM neural network model parameters until the preset target is met.
As a preferable scheme of the final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, the application comprises the following steps: calculating a residual sequence of a model prediction result and a reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and comprising:
setting N continuous residual sequences as the length of a sliding window, calculating the average value of all residual errors in the sliding window, and determining a threshold according to the maximum value of the residual error average values, wherein the threshold is expressed as:
wherein E is AN For the early warning threshold value, k is a set early warning coefficient,is the maximum value of the average value of the residual sequence in the sliding window when the equipment is in normal operation.
In a second aspect, the present application provides a VMD-EWOA-LSTM based final stage reheater steam temperature warning system, comprising,
the data acquisition processing module is used for acquiring the measuring point history data of the steam temperature of the outlet of the final-stage reheater in a normal working state and carrying out VMD variation modal decomposition on the measuring point history data;
the measuring point signal processing module is used for reconstructing and denoising the measuring point signal according to the VMD variation modal decomposition result and combining the original signal correlation degree of the measuring point to obtain a denoised historical data set;
the model building module is used for preprocessing the historical data set, building a reheater outlet steam temperature prediction model based on preprocessed measuring point data, and optimizing super parameters of the model by improving whale algorithm;
and the early warning module is used for calculating a residual sequence of the model prediction result and the reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and warning the temperature of the final-stage reheater when the average value of the final-stage reheater outlet steam temperature residual is larger than the set early warning threshold.
In a third aspect, the present application provides a computing device comprising:
a memory and a processor;
the memory is used for storing computer executable instructions, and the processor is used for executing the computer executable instructions, and the steps of the final stage reheater steam temperature early warning method based on the VMD-EWOA-LSTM are realized when the computer executable instructions are executed by the processor. In a fourth aspect, the present application provides a computer readable storage medium storing computer executable instructions that when executed by a processor implement the steps of the VMD-EWOA-LSTM based final stage reheater steam temperature warning method.
Compared with the prior art, the application has the beneficial effects that: according to the application, VMD variation modal decomposition is respectively carried out on the original signals of the measuring points of the thermal power plant, noise modal components with low correlation degree with the original signals are removed, then reconstruction denoising is carried out on the original signals, a reheater outlet steam temperature prediction model established by using the denoised measuring point data is utilized, a whale optimization algorithm is improved by introducing self-adaptive weight and variation factors, and an LSTM neural network model is combined, so that the global searching capacity of the model and the accuracy of the prediction model are improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. Wherein:
FIG. 1 is a flow chart of a method and system for final stage reheater steam temperature early warning based on VMD-EWOA-LSTM according to an embodiment of the present application;
FIG. 2 is a graph comparing a signal after denoising of a coal amount with an original signal of a final-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM according to an embodiment of the present application;
FIG. 3 is a graph comparing a primary steam pressure denoised signal with an original signal of a final stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM according to an embodiment of the present application;
FIG. 4 is a graph comparing a signal after denoising of reheat steam with an original signal of a final stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM according to an embodiment of the present application;
FIG. 5 is a graph comparing EWOA and WOA fitness curves of a final stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM according to an embodiment of the present application;
FIG. 6 is a graph of actual and predicted reheat temperatures for VMD-EWOA-LSTM in a test set, according to one embodiment of the present application, for a final stage reheater steam temperature warning method and system based on VMD-EWOA-LSTM;
FIG. 7 is a schematic diagram of a final stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM, with actual and predicted residuals of reheat steam temperature according to an embodiment of the present application;
FIG. 8 is a schematic diagram of a final stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM, and a reheat steam temperature overtemperature early warning result according to an embodiment of the present application.
Detailed Description
So that the manner in which the above recited objects, features and advantages of the present application can be understood in detail, a more particular description of the application, briefly summarized above, may be had by reference to the embodiments, some of which are illustrated in the appended drawings. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application, but the present application may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present application is not limited to the specific embodiments disclosed below.
Further, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic can be included in at least one implementation of the application. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
While the embodiments of the present application have been illustrated and described in detail in the drawings, the cross-sectional view of the device structure is not to scale in the general sense for ease of illustration, and the drawings are merely exemplary and should not be construed as limiting the scope of the application. In addition, the three-dimensional dimensions of length, width and depth should be included in actual fabrication.
Also in the description of the present application, it should be noted that the orientation or positional relationship indicated by the terms "upper, lower, inner and outer", etc. are based on the orientation or positional relationship shown in the drawings, are merely for convenience of describing the present application and simplifying the description, and do not indicate or imply that the apparatus or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present application. Furthermore, the terms "first, second, or third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
The terms "mounted, connected, and coupled" should be construed broadly in this disclosure unless otherwise specifically indicated and defined, such as: can be fixed connection, detachable connection or integral connection; it may also be a mechanical connection, an electrical connection, or a direct connection, or may be indirectly connected through an intermediate medium, or may be a communication between two elements. The specific meaning of the above terms in the present application will be understood in specific cases by those of ordinary skill in the art.
Example 1
Referring to FIG. 1, in one embodiment of the present application, a final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM is provided, comprising:
s1, acquiring measuring point historical data of steam temperature of an outlet of a final-stage reheater in a normal working state, and performing VMD variation modal decomposition on the measuring point historical data;
further, the measurement point history data at least includes: coal amount, smoke discharging temperature, total fuel amount, total air quantity, main steam pressure, opening of a secondary water spraying valve, opening of a smoke baffle plate and outlet pressure of a reheater.
Further, performing VMD-variational modal decomposition on the survey point data includes:
establishing a variation constraint model according to the acquired measurement point historical data, and introducing a penalty factor and a Lagrange operator to convert the constraint model into a non-constraint model;
specifically, the variation constraint model is expressed as:
wherein { u } k Sum { omega } k The delta (t) is a dirac function, k is the number of decomposition layers, f is the original signal,to u after Hilbert transformation k Spectrum of (t): convolution operation, +.>For gradient operation, j is the number of constraints, +.>Representing a frequency omega k Is a complex exponential signal of (a).
Specifically, the unconstrained model is expressed as:
where α is the bandwidth and λ is the lagrangian.
And solving saddle point solutions of the unconstrained model by using an overlap direction multiplier method to obtain each IMF component.
Further, solving a saddle point solution of the unconstrained model to obtain each IMF component includes:
will be{ω k 1 }、{λ 1 Initializing the number K of modes and the bandwidth alpha to 0;
for a pair ofAnd lambda (lambda) 1 Iterative updating is performed, expressed as:
wherein k is the decomposition number of layers, u k Represents the kth IMF mode function, ω k Representing the center frequency corresponding to the kth IMF mode function, f representing the original signal, and τ being the time scale of the mode function;
the iteration termination condition is expressed as:
wherein epsilon is the judgment precision;
when the iteration result is smaller than the judgment precision, outputting K modal components;
and when the iteration result is not less than the judgment precision, continuing to update the iteration.
S2, reconstructing and denoising the measuring point signals according to VMD variation modal decomposition results and combining the original signal correlation degree of the measuring point to obtain a denoised historical data set;
further, the spearman coefficient of each IMF component and each measuring point original signal is obtained through calculating VMD decomposition, and is expressed as:
wherein X is i For each IMF component obtained after VMD decomposition, Y i Is the original signal of each measuring point.
Furthermore, the reconstruction denoising of the measuring point signal by combining the original signal correlation degree of the measuring point comprises the following steps:
when the Spekerman coefficient of the IMF component and the original signal of the measuring point is smaller than 0.1, eliminating the IMF noise component;
when the Spekerman coefficient of the IMF component and the original signal of the measuring point is not less than 0.1, reserving the IMF component;
it should be noted that, after VMD decomposition of the measured point data, many IMF components are obtained, only a small part of which is noise component, and only IMF components with a spearman coefficient smaller than 0.1 with the original signal are regarded as noise components. And adding the reserved IMF components to obtain a noise-reduced historical data set.
S3, preprocessing a historical data set, establishing a reheater outlet steam temperature prediction model based on preprocessed measurement point data, and optimizing super parameters of the model by improving whale algorithm;
furthermore, the method for establishing the reheater outlet steam temperature prediction model based on the preprocessed measuring point data and optimizing the super-parameters of the model by improving the whale algorithm comprises the following steps: taking the normalized measuring point data as an input variable of a reheater outlet steam temperature prediction model;
specifically, the denoised historical data set is normalized, expressed as:
wherein x is an original signal, and x' is a normalized signal of each measuring point. .
Preferably, the model parameters of the LSTM neural network are optimized by using a whale optimization algorithm introducing self-adaptive weights and variation factors, and the model parameters of the optimized LSTM neural network at least comprise an initialization learning rate, the number of hidden layers and a regularization coefficient;
specifically, initializing parameters of an improved whale optimization algorithm EWOA at least comprises population number and maximum iteration number, and setting the dimension of the optimization parameters and the corresponding upper and lower limit ranges.
Calculating the fitness value of each individual in the population, and recording the optimal fitness value and the corresponding position vector in the current population as the optimal individual, wherein the fitness function is the root mean square error of the predicted value and the actual value of the reheat steam temperature, and the fitness function is expressed as:
wherein y is the actual value of the reheat steam temperature,is a predicted value of reheat steam temperature.
The convergence factor for each individual in the population is calculated and updated using a linear decrementing strategy, expressed as:
where T is the current iteration number, T is the maximum iteration number, and a is a linearly decreasing parameter from 2 to 0.
Introducing self-adaptive weight, wherein the weight drops slowly in the iteration starting stage, so that the algorithm has better global exploration capacity; after a certain number of iterations, the weight is rapidly reduced, so that the algorithm can search the optimal solution more finely in the local development stage, and the method is expressed as follows:
wherein ω is a weight factor.
It should be noted that, when P <0.5, |a|Σ1, whale groups randomly search by randomly selecting individual position vectors, update the next generation of positions, expressed as:
wherein, r1 and r2 are [0,1]]Is a random number of (a) and (b),is an individual randomly selected from the current population, < +.>For the current position of whale individuals, +.>For the optimal whale position at the current iteration number, < >>For the current position of a random whale +.>And->Is the distance of the current whale position from the prey.
When P <0.5, |a| <1, the whale update position encloses the prey, and the mutation factor F is introduced, so that the algorithm can jump out of local optimum more easily and increase population diversity, which is expressed as:
wherein the range of the variation factor F is [0,1].
When P + 0.5, whale screw shrinkage predatory prey, expressed as:
wherein b is a spiral constant and l is a random number of [ -1, 1].
Ending the iteration when t=t reaches the maximum iteration number and the convergence factor a is reduced from 2 to 0, otherwise continuing the loop; the output optimal whale positions respectively correspond to the number of hidden layers of the optimized LSTM, the initialization learning rate and the regularization coefficient.
Outputting a reheat steam temperature predicted value when the training result of the reheater outlet steam temperature predicted model meets a preset target; otherwise, continuing to optimize the LSTM neural network model parameters until the preset target is met.
S4, calculating a residual sequence of a model prediction result and a reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and alarming the temperature of the final reheater when the average value of the final reheater outlet steam temperature residual is larger than the set early warning threshold;
further, calculating a residual sequence of the model prediction result and the reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, including:
the fitness function of the whale optimization algorithm is utilized to be the root mean square error of the reheat steam temperature predicted value and the actual value; setting N continuous residual sequences as the length of a sliding window and calculating the average value of all residual errors in the sliding window, wherein the average value is expressed as:
wherein E is k For the residual sequence, N is the length of the sliding window,for the average value of the residual sequence in the sliding window
It should be noted that the sliding window will slide forward one residual point at a time, further calculating the average of all residuals over the window length.
Determining a threshold value according to the maximum value of the residual average value, wherein the threshold value is expressed as:
wherein E is AN For the early warning threshold value, k is a set early warning coefficient,is the maximum value of the average value of the residual sequence in the sliding window when the equipment is in normal operation.
The above is a schematic scheme of a final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM in this embodiment. It should be noted that, the technical solution of the system and the technical solution of the final-stage reheater steam temperature early warning method based on VMD-EWOA-LSTM belong to the same concept, and in this embodiment, details of the technical solution of the final-stage reheater steam temperature early warning system based on VMD-EWOA-LSTM are not described in detail, and all reference may be made to the description of the technical solution of the final-stage reheater steam temperature early warning method based on VMD-EWOA-LSTM.
In this embodiment, a final stage reheater steam temperature early warning system based on VMD-EWOA-LSTM includes:
the data acquisition processing module is used for acquiring the measuring point history data of the steam temperature of the outlet of the final-stage reheater in a normal working state and carrying out VMD variation modal decomposition on the measuring point history data;
the measuring point signal processing module is used for carrying out reconstruction denoising on the measuring point signal according to the VMD variation modal decomposition result and combining the original signal correlation degree of the measuring point to obtain a denoised historical data set;
the model building module is used for preprocessing the historical data set, building a reheater outlet steam temperature prediction model based on preprocessed measuring point data, and optimizing super parameters of the model by improving whale algorithm;
and the early warning module is used for calculating a residual sequence of a model prediction result and a reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and warning the temperature of the final reheater when the average value of the final reheater outlet steam temperature residual is larger than the set early warning threshold.
The embodiment also provides a computing device, which is suitable for the situation of a final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM, and comprises the following steps:
a memory and a processor; the memory is used for storing computer executable instructions, and the processor is used for executing the computer executable instructions to realize the final stage reheater steam temperature early warning method based on the VMD-EWOA-LSTM as proposed by the embodiment.
The present embodiment also provides a storage medium having stored thereon a computer program which, when executed by a processor, implements a final stage reheater steam temperature warning method based on VMD-EWOA-LSTM as proposed in the above embodiment.
The storage medium proposed in this embodiment belongs to the same inventive concept as the last-stage reheater steam temperature early warning method based on VMD-EWOA-LSTM proposed in the above embodiment, and technical details not described in detail in this embodiment can be seen in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.
From the above description of embodiments, it will be clear to a person skilled in the art that the present application may be implemented by means of software and necessary general purpose hardware, but of course also by means of hardware, although in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as a floppy disk, a Read Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, etc., including several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to execute the method of the embodiment of the present application.
Example 2
Referring to fig. 2 to 8, for one embodiment of the present application, the beneficial effects of my application were verified through specific experiments.
And selecting data of each measuring point which is related to the steam temperature of the outlet of the final-stage reheater and comprises 2690 measuring points such as coal amount of a coal feeder, smoke exhaust temperature, total fuel amount, total air quantity, main steam pressure, opening of a secondary water spraying valve, opening of a smoke baffle, outlet pressure of the reheater and the like, and acquiring the data once every 5 minutes, wherein the first 90 percent of data are a normal operation data set, and the last 10 percent of data are fault data sets.
Performing VMD variation modal decomposition on 2421 data in the normal operation data set of each measuring point, and setting a decomposition layer number K=7 and a penalty factor alpha=2000; and respectively calculating the Speerman coefficient of each IMF component obtained by VMD decomposition and each measuring point original signal, removing the noise component with the absolute value of the Speerman coefficient of each measuring point original signal lower than 0.1, and adding the corresponding values of the rest modal components to obtain each measuring point signal after noise reduction.
Referring to fig. 2 to 4, a comparison graph of the signals after denoising of the three measuring points of the coal amount, the main steam pressure and the reheat steam pressure of the coal feeder and the original signals is respectively listed, and it can be seen that the denoising signals reconstructed after VMD decomposition can effectively remove noise information in the original signals and also can remove abnormal values in the original signals.
Normalizing the data set of each measuring point after noise reduction, taking the preprocessed normal operation data of each measuring point as an input variable, and utilizing an improved whale optimization algorithm EWOA to optimize LSTM related parameters including an initialization learning rate, the number of hidden layers and a regularization coefficient, thereby establishing a final-stage reheater outlet steam temperature prediction model.
The first 80% of the total 1956 data of the normal operation data of each measuring point are divided into training sets, and the last 20% of the total 465 data are divided into test sets. Initializing relevant parameters of an EWOA improved whale optimization algorithm, setting the population number to 30, setting the maximum iteration number to 40, and setting the dimension of the optimization parameters to 3. The optimizing range of the initialization learning rate is [0.001,0.15], the optimizing range of the number of hidden layers is [1,100], and the optimizing range of the regularization coefficient is [0.1,0.3].
Calculating fitness value of each individual in the population, recording optimal fitness value and corresponding position vector in the current population as optimal individual, wherein fitness function is root mean square error of reheat steam temperature predicted value and actual value in training set, calculating and updating convergence factor a of each individual in the population by using linear decreasing strategy, a is a linearly decreasing parameter from 2 to 0,
as can be seen from fig. 5, compared with WOA, the EWOA overcomes the disadvantages of the conventional whale optimization algorithm, such as poor global searching capability and easy sinking into local optimization, by introducing the adaptive weight ω and the variation factor F, and the optimal parameters of LSTM obtained after EWOA optimization include an initialization learning rate of 0.095, a number of hidden layers of 11, and a regularization coefficient of 0.3.
FIG. 6 is a graph of predicted and actual reheat steam temperature values of the VMD-EWOA-LSTM model in a test set. The prediction evaluation results of the embodiment model of the application in the test set are compared with those of other models, as shown in table 1:
table 1 comparison of model evaluation results
| Model | R 2 | MSE | RMSE |
| PSO-LSTM | 0.784 | 2.766 | 1.663 |
| WOA-LSTM | 0.843 | 2.161 | 1.47 |
| EWOA-LSTM | 0.894 | 1.839 | 1.356 |
| VMD-EWOA-LSTM | 0.942 | 1.66 | 1.288 |
Residual sequence E of reheat steam temperature predicted value and actual operation value in VMD-EWOA-LSTM model is calculated k Setting the width n=10 of the sliding window, and calculating the average value of the residual sequence according to the following formula; as can be seen from fig. 7, when the working condition is normal, the residual error is evenly distributed between (-0.3, 0.3), and the early warning coefficient k=1.1 is set to obtain the upper and lower early warning thresholds E of the reheat steam temperature AN 0.33.
As can be seen from the early warning result of FIG. 8, when the reheat steam temperature is normal, the average value of the residual error is distributed between the upper and lower thresholds, and when the reheat steam temperature is in a degradation trend, the average value of the residual error is gradually increased until the threshold is exceeded, in FIG. 8, the 409 th measuring point exceeds the threshold to generate early warning, and the actual fault occurs at the 415 th measuring point. Therefore, the method can accurately predict the abnormal state of the reheat steam temperature by 6 measuring points in advance, namely 30 minutes in advance, and strives for time for timely controlling the reheat steam temperature.
It should be noted that the above embodiments are only for illustrating the technical solution of the present application and not for limiting the same, and although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present application may be modified or substituted without departing from the spirit and scope of the technical solution of the present application, which is intended to be covered in the scope of the claims of the present application.
Claims (10)
1. The final stage reheater steam temperature early warning method based on VMD-EWOA-LSTM is characterized by comprising the following steps:
acquiring measuring point historical data of steam temperature of an outlet of a final-stage reheater in a normal working state, and performing VMD variation modal decomposition on the measuring point historical data;
reconstructing and denoising the measuring point signals according to the VMD variation modal decomposition result and combining the original signal correlation degree of the measuring point to obtain a denoised historical data set;
preprocessing the historical data set, establishing a reheater outlet steam temperature prediction model based on preprocessed measuring point data, and optimizing super parameters of the model by improving whale algorithm;
and calculating a residual sequence of the model prediction result and the reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and alarming the temperature of the final-stage reheater when the average value of the final-stage reheater outlet steam temperature residual is larger than the set early warning threshold.
2. The final stage reheater steam temperature warning method based on VMD-EWOA-LSTM of claim 1, wherein said station history data includes at least: coal amount of a coal feeder, smoke discharging temperature, total fuel amount, total air quantity, main steam pressure, opening of a secondary water spraying valve, opening of a smoke baffle plate and outlet pressure of a reheater.
3. The final stage reheater steam temperature warning method based on VMD-EWOA-LSTM of claim 2, wherein performing VMD variation modal decomposition on the measurement point data comprises:
establishing a variation constraint model according to the acquired measurement point historical data, and introducing a penalty factor and a Lagrange operator to convert the constraint model into a non-constraint model;
and solving saddle point solutions of the unconstrained model by using an overlap direction multiplier method to obtain each IMF component.
4. The final-stage reheater steam temperature early warning method based on VMD-EWOA-LSTM according to claim 3, wherein the solving the saddle point solution of the unconstrained model to obtain each IMF component includes:
will be{ω k 1 }、{λ 1 Initializing the number K of modes and the bandwidth alpha to 0;
for a pair ofAnd lambda (lambda) 1 Iterative updating is performed, expressed as:
where k is the number of decomposition layers, α is the bandwidth, λ is the Lagrangian operator, { u k The k-th IMF mode function, { ω k The center frequency corresponding to the kth IMF mode function is represented, f represents the original signal, and tau is the time scale of the mode function;
the iteration termination condition is expressed as:
wherein epsilon is the judgment precision;
when the iteration result is smaller than the judgment precision, outputting K modal components;
and when the iteration result is not less than the judgment precision, continuing to update the iteration.
5. The final stage reheater steam temperature warning method based on VMD-EWOA-LSTM of claim 4, wherein the reconstruction denoising of the measurement point signal by combining the measurement point original signal correlation comprises:
when the Spekerman coefficient of the IMF component and the original signal of the measuring point is smaller than 0.1, eliminating the IMF noise component;
when the Spekerman coefficient of the IMF component and the original signal of the measuring point is not less than 0.1, reserving the IMF component;
and adding the reserved IMF components to obtain a noise-reduced historical data set.
6. The final stage reheater steam temperature warning method based on VMD-EWOA-LSTM according to claim 5, wherein the step of establishing a reheater outlet steam temperature prediction model based on the preprocessed measurement point data, and optimizing super parameters of the model by improving whale algorithm comprises: taking the normalized measuring point data as an input variable of a reheater outlet steam temperature prediction model;
optimizing LSTM neural network model parameters by using whale optimization algorithm with self-adaptive weight and variation factor, wherein the model parameters of the optimized LSTM neural network at least comprise an initialization learning rate, the number of hidden layers and a regularization coefficient;
the fitness function of the whale optimization algorithm is root mean square error of reheat steam temperature predicted value and actual value
Outputting a reheat steam temperature predicted value when the training result of the reheater outlet steam temperature predicted model meets a preset target; otherwise, continuing to optimize the LSTM neural network model parameters until the preset target is met.
7. The final-stage reheater steam temperature early warning method based on VMD-EWOA-LSTM according to claim 6, wherein calculating a residual sequence of a model prediction result and a reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, comprises:
setting N continuous residual sequences as the length of a sliding window, calculating the average value of all residual errors in the sliding window, and determining a threshold according to the maximum value of the residual error average values, wherein the threshold is expressed as:
wherein E is AN For the early warning threshold value, k is a set early warning coefficient,is the maximum value of the average value of the residual sequence in the sliding window when the equipment is in normal operation.
8. The final stage reheater steam temperature early warning system based on VMD-EWOA-LSTM is characterized in that the system comprises:
the data acquisition processing module is used for acquiring the measuring point history data of the steam temperature of the outlet of the final-stage reheater in a normal working state and carrying out VMD variation modal decomposition on the measuring point history data;
the measuring point signal processing module is used for reconstructing and denoising the measuring point signal according to the VMD variation modal decomposition result and combining the original signal correlation degree of the measuring point to obtain a denoised historical data set;
the model building module is used for preprocessing the historical data set, building a reheater outlet steam temperature prediction model based on preprocessed measuring point data, and optimizing super parameters of the model by improving whale algorithm;
and the early warning module is used for calculating a residual sequence of the model prediction result and the reheat steam temperature actual value, setting an early warning threshold according to the residual sequence, and warning the temperature of the final-stage reheater when the average value of the final-stage reheater outlet steam temperature residual is larger than the set early warning threshold.
9. An electronic device, comprising:
a memory and a processor;
the memory is configured to store computer executable instructions that, when executed by the processor, implement the steps of the VMD-EWOA-LSTM based final stage reheater steam temperature warning method of any one of claims 1 to 7.
10. A computer readable storage medium storing computer executable instructions which when executed by a processor implement the steps of the VMD-EWOA-LSTM based final stage reheater steam temperature warning method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310847975.7A CN117093836A (en) | 2023-07-11 | 2023-07-11 | Last-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310847975.7A CN117093836A (en) | 2023-07-11 | 2023-07-11 | Last-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117093836A true CN117093836A (en) | 2023-11-21 |
Family
ID=88772515
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310847975.7A Pending CN117093836A (en) | 2023-07-11 | 2023-07-11 | Last-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117093836A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117648645A (en) * | 2023-12-08 | 2024-03-05 | 国能宁夏鸳鸯湖第一发电有限公司 | Main steam temperature prediction method based on Bayes-Catboost |
-
2023
- 2023-07-11 CN CN202310847975.7A patent/CN117093836A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117648645A (en) * | 2023-12-08 | 2024-03-05 | 国能宁夏鸳鸯湖第一发电有限公司 | Main steam temperature prediction method based on Bayes-Catboost |
| CN117648645B (en) * | 2023-12-08 | 2024-06-11 | 国能宁夏鸳鸯湖第一发电有限公司 | Main steam temperature prediction method based on Bayes-Catboost |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117093836A (en) | Last-stage reheater steam temperature early warning method and system based on VMD-EWOA-LSTM | |
| Ebrahimgol et al. | A novel approach in exergy optimization of a WWER1000 nuclear power plant using whale optimization algorithm | |
| CN110570030A (en) | Wind power cluster power interval prediction method and system based on deep learning | |
| CN112884237A (en) | Power distribution network prediction auxiliary state estimation method and system | |
| CN113222289B (en) | Prediction method of energy power based on data processing | |
| CN110989342B (en) | Real-time T-S fuzzy modeling method for combined cycle unit heavy-duty gas turbine | |
| CN112052570A (en) | Economy backpressure optimization method of wet cooling unit of thermal power plant based on wolf algorithm | |
| CN117117859B (en) | Photovoltaic power generation power prediction method and system based on neural network | |
| CN112377985B (en) | Heat supply unit peak regulation lower limit evaluation method and system based on feedback neural network | |
| CN112417764A (en) | K nearest neighbor regression prediction method for boiler special equipment steam flow prediction | |
| CN111144663A (en) | Ultra-short-term wind power prediction method for offshore wind farm considering output fluctuation process | |
| CN112328971B (en) | Random number-based bimodal three-parameter Weibull distribution fitting method | |
| CN116840720A (en) | Fuel cell remaining life prediction method | |
| CN116933922A (en) | WOA-CNN-LSTM-based photovoltaic power generation power prediction method | |
| CN111931395A (en) | Sensor measuring point optimization method for reducing strain field reconstruction errors | |
| CN116578551A (en) | GRU-GAN-based power grid data restoration method | |
| CN114818925A (en) | Method and system for filling missing value of output data of wind power plant | |
| CN115345297A (en) | Platform area sample generation method and system based on generation countermeasure network | |
| Sindareh-Esfahani et al. | Model predictive control of a heat recovery steam generator during cold start-up operation using piecewise linear models | |
| CN113222263A (en) | Photovoltaic power generation power prediction method based on long-term and short-term memory neural network | |
| CN114819382B (en) | LSTM-based photovoltaic power prediction method | |
| CN113836819B (en) | Bed temperature prediction method based on time sequence attention | |
| CN115217152A (en) | Method and device for predicting opening and closing deformation of immersed tunnel pipe joint | |
| CN115498627A (en) | Combined load prediction method integrating sequence decomposition and feature selection | |
| CN113536508A (en) | Method and system for classifying manufacturing network nodes |
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 |