CN114759613A - LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system - Google Patents

LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system Download PDF

Info

Publication number
CN114759613A
CN114759613A CN202210550300.1A CN202210550300A CN114759613A CN 114759613 A CN114759613 A CN 114759613A CN 202210550300 A CN202210550300 A CN 202210550300A CN 114759613 A CN114759613 A CN 114759613A
Authority
CN
China
Prior art keywords
network
sequence
time
lstm
time sequence
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
CN202210550300.1A
Other languages
Chinese (zh)
Inventor
张小科
胡怀中
王子杰
夏大伟
王景钢
曹桂州
李珍平
史书怀
张步庭
李玲
陈二强
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.)
Xian Jiaotong University
Electric Power Research Institute of State Grid Henan Electric Power Co Ltd
Original Assignee
Xian Jiaotong University
Electric Power Research Institute of State Grid Henan Electric Power 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 Xian Jiaotong University, Electric Power Research Institute of State Grid Henan Electric Power Co Ltd filed Critical Xian Jiaotong University
Priority to CN202210550300.1A priority Critical patent/CN114759613A/en
Publication of CN114759613A publication Critical patent/CN114759613A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
    • H02J3/466—Scheduling the operation of the generators, e.g. connecting or disconnecting generators to meet a given demand
    • 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
    • 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
    • G06N3/048—Activation functions
    • 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
    • G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
    • 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
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Biophysics (AREA)
  • Evolutionary Computation (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Mathematical Physics (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Power Engineering (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • Primary Health Care (AREA)
  • Control Of Turbines (AREA)

Abstract

The invention discloses an LSTM-based online estimation method and system for primary frequency modulation capability of a deep peak-shaving thermal power generating unit. The method is suitable for the deep peak regulation working condition, has higher prediction accuracy on the primary frequency modulation load of the thermal power generating unit, and is beneficial to fully exerting the primary frequency modulation capability of the deep peak regulation thermal power generating unit.

Description

LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system
Technical Field
The invention belongs to the technical field of primary frequency modulation of thermal power generating units, and particularly relates to an LSTM-based online estimation method and system for primary frequency modulation capability of a deep peak shaving unit.
Background
With the rapid development of new energy power generation in China, the proportion of the traditional coal-fired unit in a power grid is smaller and smaller. The clean energy power generation has a series of advantages which are not possessed by the traditional thermal power generation, but the fluctuation and intermittence of the clean energy power generation output can also cause the instability of the power grid frequency, so that new requirements on the peak regulation and frequency modulation capability of the power grid are provided. In the future, for a long time, the thermal power generating unit can undertake the main tasks of power grid peak regulation and frequency modulation, and increasingly participate in deep peak regulation operation of the power grid. In order to ensure safe and stable operation of a power grid and fully exert the primary frequency modulation capability of a deep peaking generation thermal power unit, it is necessary to research a primary frequency modulation load online prediction method of the thermal power unit during the deep peaking generation operation.
The existing method for predicting the primary frequency modulation load of the thermal power generating unit is mainly realized by mechanism modeling for constructing a thermal power generating unit coordinated control system and a digital electro-hydraulic control system, but the system of the thermal power generating unit is highly complex, equipment mainly related to a primary frequency modulation process of the thermal power generating unit comprises a steam turbine, a reheater, a digital electro-hydraulic system and the like, the change of main steam parameters of the thermal power generating unit and the flow characteristic of a valve of the steam turbine have important influence on the primary frequency modulation capacity, the operation characteristic of the thermal power generating unit is greatly changed due to deep peak modulation operation, the complex characteristic of the primary frequency modulation of the deep peak modulation thermal power generating unit is difficult to be fully described by the method for predicting through the mechanism modeling, massive historical data of the operation of the thermal power generating unit are not fully utilized, and the prediction accuracy is poor. Therefore, in order to fully exert the primary frequency modulation capability of the deep peaking thermal power generating unit and ensure the safety of a power grid, a more accurate online prediction method for the primary frequency modulation load of the deep peaking thermal power generating unit is needed.
Disclosure of Invention
The invention aims to solve the technical problem that the method and the system for estimating the primary frequency modulation capacity of the deep peak-shaving thermal power generating unit on line based on the LSTM are provided aiming at the defects in the prior art, and the primary frequency modulation characteristic of equipment and links influencing the primary frequency modulation of the thermal power generating unit under the deep peak-shaving working condition is captured by utilizing the unique advantages of the LSTM deep neural network in processing the time sequence problem, so that the online prediction of the primary frequency modulation load of the deep peak-shaving thermal power generating unit is realized.
The invention adopts the following technical scheme:
the LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method comprises the following steps:
respectively constructing a boiler combustion system sub-network based on an LSTM neural network, a reheating system sub-network based on the LSTM neural network, a steam turbine speed regulating system sub-network based on an MLP (Multi-level programmable processor) and a steam turbine sub-network based on the LSTM neural network, constructing a time sequence input sequence and a time sequence target sequence with uniform time dimension according to the operation historical data of the deep peak-shaving thermal power generating unit, and respectively training the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulating system sub-network and the steam turbine sub-network by utilizing the corresponding time sequence input sequence and the time sequence target sequence;
Combining the trained boiler combustion system sub-network, reheating system sub-network, steam turbine speed regulation system sub-network and steam turbine sub-network to obtain a complete primary frequency regulation load prediction network of the deep peaking thermal power generating unit;
inputting the operation historical data of the deep peaking generation unit into the obtained primary frequency modulation load prediction network of the deep peaking generation unit, and performing short-term prediction on main steam pressure, main steam temperature, pressure after valve adjustment, reheat steam pressure, reheat steam temperature and unit load of a future unit;
after the deep peak-shaving thermal power generating unit obtains new sampling data, the new sampling data is added into the input data, data with the same time length at the initial moment are removed from the original input data, the network state is updated, short-term prediction is repeated, and primary frequency modulation load on-line prediction of the deep peak-shaving thermal power generating unit is achieved.
Specifically, the construction of the boiler combustion system sub-network based on the LSTM neural network specifically includes:
constructing a fuel quantity, a valve regulating instruction, a water supply quantity and a air supply quantity in operation historical data of a deep peaked thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing main steam pressure and main steam temperature in the operation historical data of the deep peaked thermal power generating unit into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the time sequence target sequence have the same time length, the time sequence target sequence lags behind the time sequence input sequence, and the lag time length is the time length to be predicted;
Constructing a boiler combustion system sub-network based on an LSTM neural network, standardizing the constructed time sequence input sequence and the time sequence target sequence by adopting a standard deviation method, and training the boiler combustion system sub-network based on the LSTM by using the standardized time sequence input sequence and the standardized time sequence target sequence to obtain the trained boiler combustion system sub-network.
Specifically, constructing a reheating system sub-network based on the LSTM neural network specifically includes:
constructing fuel quantity, valve regulating instruction, water supply quantity and air supply quantity in the operation historical data of the deep peaked thermal power generating unit into a time sequence input sequence with uniform time dimension; constructing the reheat steam pressure and the reheat steam temperature in the operation historical data of the deep peaked thermal power generating unit into a time sequence target sequence with uniform time dimension; the time sequence input sequence and the time sequence target sequence have the same time length, the time sequence target sequence lags behind the time sequence input sequence, and the lag time length is the time length to be predicted;
constructing a reheating system sub-network based on an LSTM neural network, standardizing the constructed time sequence input sequence and the time sequence target sequence by adopting a standard deviation method, and training the reheating system sub-network based on the LSTM by using the time sequence input sequence and the time sequence target sequence after the standardization so as to obtain the trained reheating system sub-network.
Specifically, the construction of the steam turbine speed regulating system sub-network specifically comprises the following steps:
constructing a valve regulating instruction, main steam pressure and main steam temperature in operation historical data of a deep peaked thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing regulating stage pressure in the operation historical data of the deep peaked thermal power generating unit into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the time sequence target sequence have the same time length, and the time of the time sequence input sequence corresponds to the time of the time sequence target sequence one by one;
and constructing a steam turbine speed regulating system sub-network, standardizing the constructed time sequence input sequence and the time sequence target sequence by adopting a standard deviation method, and training the steam turbine speed regulating system sub-network by using the standardized time sequence input sequence and the standardized time sequence target sequence to obtain the trained steam turbine speed regulating system sub-network.
Furthermore, the sub-network of the speed regulating system of the steam turbine comprises a sequence input layer, a plurality of full connection layers, a plurality of Drop layers and a sequence output layer, wherein the full connection layer Yt＝WXt+ b, W tableIndicating fully connected layer learnable weights, b indicating learnable bias weights, XtIndicating the input of the fully-connected layer at time t, Y tIndicating the output of the fully connected layer at time t.
Specifically, the construction of the steam turbine sub-network based on the LSTM neural network specifically includes:
constructing a regulating stage pressure, a reheat steam pressure and a reheat steam temperature in the operation historical data of the deep peaking thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing a unit load in the operation historical data of the deep peaking thermal power generating unit into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the time sequence target sequence have the same time length, and the time of the time sequence input sequence corresponds to the time of the time sequence target sequence one by one;
and constructing a steam turbine sub-network based on the LSTM neural network, standardizing the constructed time sequence input sequence and the target sequence by adopting a standard deviation method, and training the steam turbine sub-network based on the LSTM by using the standardized time sequence input sequence and the standardized time sequence target sequence to obtain the trained steam turbine sub-network.
Specifically, the time sequence input sequence with uniform time dimension is:
Figure BDA0003654746700000041
the timing target sequence with uniform time dimension is:
Figure BDA0003654746700000042
N1、N2the characteristic dimensions of data types in the input sequence and the target sequence are respectively represented, T represents a time dimension, and the time sequence input sequence comprises original data values with T time lengths and data change values with T time lengths.
Specifically, the boiler combustion system sub-network based on the LSTM neural network, the reheating system sub-network based on the LSTM neural network and the steam turbine sub-network based on the LSTM neural network respectively comprise a sequence input layer, a plurality of LSTM layers, a plurality of full connection layers, a plurality of Drop layers and a sequence output layer.
Further, the LSTM memory cells of the LSTM layer are calculated as follows:
Figure BDA0003654746700000043
ct＝ft⊙ct-1+it⊙gt
ht＝ot⊙σc(ct)
it＝σg(Wixt+Riht-1+bi)
ft＝σg(Wfxt+Rfht-1+bf)
gt＝σc(Wgxt+Rght-1+bg)
ot＝σg(Woxt+Roht-1+bo)
σ(x)＝(1+e-x)-1
wherein W represents a learnable input weight, R represents a learnable cyclic weight, b represents a learnable bias; i. f, g and o respectively represent an input gate, a forgetting gate, a candidate unit and an output gate; c. CtRepresenting the state of the cell at time step t; h is a total oftRepresenting the hidden state at time step t; sigmacRepresenting the tan h state activation function; sigmagRepresenting a sigmoid gate activation function; full connection layer Yt＝WXt+ b, W represents the fully connected layer learnable weight, b represents the learnable bias weight, XtIndicating the input of the fully-connected layer at time t, YtRepresenting the output of the fully connected layer at time t.
In a second aspect, an embodiment of the present invention provides an online estimation system for a primary frequency modulation capability of a depth peak shaver set based on LSTM, including:
the network training module is used for respectively constructing a boiler combustion system sub-network based on an LSTM neural network, a reheating system sub-network based on the LSTM neural network, a steam turbine speed regulation system sub-network and a steam turbine sub-network based on the LSTM neural network, constructing a time sequence input sequence and a time sequence target sequence with uniform time dimension according to the operation historical data of the deep peak regulation thermal power generating unit, and respectively training the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network by utilizing the corresponding time sequence input sequence and the time sequence target sequence;
The network combination module is used for combining the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network which are trained by the network training module to obtain a complete primary frequency modulation load prediction network of the deep peaking thermal power generating unit;
the short-term prediction module is used for inputting the operation historical data of the deep peak shaving thermal power generating unit into a primary frequency modulation load prediction network of the deep peak shaving thermal power generating unit obtained by the network combination module, and performing short-term prediction on the main steam pressure, the main steam temperature, the reheated steam pressure, the reheated steam temperature, the regulation stage pressure and the unit load of a future unit;
and the online prediction module is used for adding the new sampling data into the input data after the unit acquires the new sampling data, removing the data with the same time length at the initial moment from the original input data, updating the network state, and repeating the short-term prediction module to realize the online prediction of the primary frequency modulation load of the deep peaking thermal power unit.
Compared with the prior art, the invention at least has the following beneficial effects:
the invention relates to an LSTM-based online estimation method for primary frequency modulation capability of a deep peak-shaving thermal power generating unit, which captures primary frequency modulation characteristics of equipment and links influencing primary frequency modulation of the thermal power generating unit under a deep peak-shaving working condition by utilizing the unique advantages of an LSTM deep neural network in processing a time sequence problem, respectively constructs a boiler combustion system sub-network, a reheating system sub-network, a steam turbine speed regulation system sub-network and a steam turbine working sub-network aiming at the corresponding links, further forms a complete model for predicting primary frequency modulation load of the deep peak-shaving thermal power generating unit, utilizes unit operation historical data to perform offline training, and utilizes unit operation real-time data to perform online prediction and state update, thereby realizing online prediction of the primary frequency modulation load of the deep peak-shaving thermal power generating unit; compared with the traditional technology, all equipment and links influencing primary frequency modulation of the unit are considered, information contained in historical operation data of the unit is fully mined, and prediction accuracy is remarkably improved.
Furthermore, the boiler combustion system sub-network based on the LSTM neural network can accurately describe the time sequence characteristics such as delay, inertia and the like expressed in the boiler combustion process, and extract the unit dynamic change information contained in the unit operation historical data, so that the estimation result of the main steam pressure has higher precision.
Furthermore, the reheating system sub-network based on the LSTM neural network can accurately describe the time sequence characteristics of the reheating system heat exchange process, and extract the unit dynamic change information contained in the unit operation historical data, so that the estimation result of the reheating steam pressure has higher precision.
Furthermore, the MLP-based steam turbine speed regulating system sub-network models the valve flow nonlinear characteristic which is difficult to describe through a mechanism through a plurality of layers of full connection layers, and the error generated by the valve flow nonlinear characteristic in the primary frequency modulation capability estimation process is eliminated.
Furthermore, the steam turbine sub-network based on the LSTM neural network can accurately describe the inertia characteristic of the steam in the steam turbine for work, and the estimation precision of the unit load is improved.
Furthermore, the time sequence input sequence comprises original historical data and a historical data change value, the setting solves the problem that the proportion of the steady-state data volume and the dynamic data volume in the sample data is maladjusted, the influence of the dynamic data is amplified, and the estimation accuracy is improved.
It is to be understood that, the beneficial effects of the second aspect may refer to the relevant description in the first aspect, and are not described herein again.
In conclusion, the method is suitable for the deep peak regulation working condition, has higher prediction accuracy on the primary frequency modulation load of the thermal power generating unit, and is beneficial to fully exerting the primary frequency modulation capability of the deep peak regulation thermal power generating unit.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
FIG. 1 is a flow chart of the operation of the present invention;
FIG. 2 is a view showing the structure of an LSTM unit employed in the present invention;
FIG. 3 is a diagram of a primary frequency modulation load prediction network structure of a complete deep peaking thermal power generating unit constructed by the invention;
fig. 4 is a diagram of simulation experiment data for primary frequency modulation capability estimation according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
In the description of the present invention, it should be understood that the terms "comprises" and/or "comprising" indicate the presence of the described features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in this specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and including such combinations, e.g., a and/or B, may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter related objects are in an "or" relationship.
It should be understood that although the terms first, second, third, etc. may be used to describe preset ranges, etc. in embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from one another. For example, a first preset range may also be referred to as a second preset range, and similarly, a second preset range may also be referred to as a first preset range, without departing from the scope of embodiments of the present invention.
The word "if," as used herein, may be interpreted as "at … …" or "when … …" or "in response to a determination" or "in response to a detection," depending on the context. Similarly, the phrase "if determined" or "if detected (a stated condition or event)" may be interpreted as "upon determining" or "in response to determining" or "upon detecting (a stated condition or event)" or "in response to detecting (a stated condition or event)", depending on the context.
Various structural schematics according to the disclosed embodiments of the invention are shown in the drawings. The figures are not drawn to scale, wherein certain details are exaggerated and possibly omitted for clarity of presentation. The shapes of various regions, layers and their relative sizes and positional relationships shown in the drawings are merely exemplary, and deviations may occur in practice due to manufacturing tolerances or technical limitations, and a person skilled in the art may additionally design regions/layers having different shapes, sizes, relative positions, according to actual needs.
The invention provides an LSTM-based online estimation method for primary frequency modulation capability of a deep peak-regulating thermal power generating unit, which captures primary frequency modulation characteristics of equipment and links influencing primary frequency modulation of the thermal power generating unit under a deep peak-regulating working condition by utilizing unique advantages of an LSTM deep neural network in processing a time sequence problem, respectively constructs a boiler combustion system sub-network, a reheating system sub-network, a steam turbine speed-regulating system sub-network and a steam turbine working sub-network aiming at corresponding links, further forms a complete model for predicting primary frequency modulation load of the deep peak-regulating thermal power generating unit, performs offline training by utilizing unit operation historical data, performs online prediction and state updating by utilizing unit operation real-time data, and realizes online prediction of the primary frequency modulation load of the deep peak-regulating thermal power generating unit. Compared with the traditional technology, the method considers all equipment and links influencing the primary frequency modulation of the unit, fully excavates the information contained in the historical operation data of the unit, and obviously improves the prediction precision.
Referring to fig. 1, the method for online estimating the primary frequency modulation capability of a depth peak shaver set based on LSTM of the present invention includes the following steps:
s1, constructing the fuel quantity, the valve regulating instruction, the water supply quantity and the air supply quantity in the operation historical data of the deep peaking thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing the main steam pressure and the main steam temperature into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence with the uniform time dimension and the target sequence have the same time length, the target sequence lags behind the input sequence, and the lag time length is the time length to be predicted;
The data sampling time is 1 second, the lag time of the target sequence compared with the input sequence is 15 seconds, and the time sequence input sequence with uniform time dimension is as follows:
Figure BDA0003654746700000091
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003654746700000092
representing the amount of fuel at the moment of the unit t,
Figure BDA0003654746700000093
indicating the set valve command at time t, wtRepresenting the feed water quantity of the unit at time t, ftThe air supply quantity of the unit at the moment t is shown,
Figure BDA0003654746700000094
representing the variation of the fuel quantity at the moment t of the unit,
Figure BDA0003654746700000095
representing the variation, Δ w, of the valve-regulating command of the unit at time tt＝wt-wt-1Representing the variation, Δ f, of the feed water quantity of the unit at time tt＝ft-ft-1And the variable quantity of the air supply quantity of the unit at the moment t is shown. In the running process of the unit, the unit runs in a stable state for most of time, the proportion of the time of the primary frequency modulation dynamic process is small, so that the proportion of dynamic data and stable data in sample data is not adjusted, and in order to solve the problem, the variation of the running parameters of the unit is also used as input data, so that the stable running data is weakened, and the influence of the dynamic data is enhanced.
The timing target sequence with a uniform time dimension is:
Figure BDA0003654746700000096
wherein the content of the first and second substances,
Figure BDA0003654746700000097
representing the main steam pressure at the moment t of the unit,
Figure BDA0003654746700000098
representing the main steam temperature at the time t of the unit.
S2, standardizing the time sequence input sequence and the target sequence with the unified time dimension in the step S1 by adopting a standard deviation method, constructing an LSTM-based boiler combustion system sub-network, and training on the basis of the standardized time sequence input sequence and the target sequence to obtain the trained boiler combustion system sub-network;
The LSTM-based boiler combustion system sub-network consists of 1 sequence input layer, 30 LSTM layers, 3 full-connection layers, 1 Drop layer and 1 sequence output layer. The LSTM unit structure in the boiler combustion system sub-network based on the LSTM deep neural network is shown in figure 2.
The standard deviation normalization method was:
Figure BDA0003654746700000101
where μ is the mean of all sample data, σ is the standard deviation of all sample data, x is*To normalize the data.
S3, constructing fuel quantity, valve regulating instruction, water supply quantity and air supply quantity in the operation historical data of the deep peaker thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing reheat steam pressure and reheat steam temperature into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence with the uniform time dimension and the target sequence have the same time length, the target sequence lags behind the input sequence, and the lag time length is the time length to be predicted;
the data sampling time is 1 second, the lag time of the target sequence compared with the input sequence is 15 seconds, and the time sequence input sequence with the uniform time dimension is as follows:
Figure BDA0003654746700000102
the timing target sequence with a uniform time dimension is:
Figure BDA0003654746700000103
wherein the content of the first and second substances,
Figure BDA0003654746700000104
representing the reheat steam pressure at time t of the unit,
Figure BDA0003654746700000105
And represents the reheat steam temperature at the moment t of the unit.
S4, standardizing the time sequence input sequence and the target sequence with the unified time dimension in the step S3 by adopting a standard deviation method, constructing an LSTM-based reheating system sub-network, and training the time sequence input sequence and the target sequence based on the standardization to obtain a trained reheating system sub-network;
the reheating system sub-network based on the LSTM deep neural network consists of 1 sequence input layer, 30 LSTM layers, 3 full-connection layers, 1 Drop layer and 1 sequence output layer. The structure of the LSTM units in the LSTM-based reheat system sub-network is shown in fig. 2. The calculation method of the standard deviation normalization is the same as the calculation method in step S2.
S5, constructing a valve regulation instruction, main steam pressure and main steam temperature in the operation historical data of the deep peaker thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing regulation level pressure into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the target sequence with the uniform time dimension have the same time length, and the time of the input sequence corresponds to the time of the target sequence one by one;
the data sampling time is 1 second, and the time sequence input sequence with the unified time dimension is as follows:
Figure BDA0003654746700000111
The timing target sequence with uniform time dimension is:
Figure BDA0003654746700000112
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003654746700000113
indicating the regulating stage pressure at the moment t of the unit.
S6, standardizing the time sequence input sequence and the target sequence with the unified time dimension in the step S5 by adopting a standard deviation method, constructing a steam turbine speed regulating system sub-network, and training based on the standardized time sequence input sequence and the target sequence to obtain the trained steam turbine speed regulating system sub-network;
in this embodiment, the steam turbine speed regulating system sub-network is composed of 1 sequence input layer, 6 full connection layers, 1 Drop layer and 1 sequence output layer. The calculation method of the standard deviation normalization is the same as the calculation method in step S2.
S7, constructing the pressure of an adjusting level, the pressure of reheated steam and the temperature of reheated steam in the operation historical data of the deep peaking thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing the load of the thermal power generating unit into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the target sequence with the uniform time dimension have the same time length, and the time of the input sequence corresponds to the time of the target sequence one by one;
the data sampling time is 1 second, and the time sequence input sequence with the unified time dimension is as follows:
Figure BDA0003654746700000121
The timing target sequence with uniform time dimension is:
Figure BDA0003654746700000122
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003654746700000123
and represents the unit load at the t moment of the unit.
S8, standardizing the time sequence input sequence and the target sequence with the unified time dimension in the S7 by adopting a standard deviation method, constructing a steam turbine sub-network based on the LSTM, and training the steam turbine sub-network based on the standardized time sequence input sequence and the target sequence to obtain the trained steam turbine sub-network;
the LSTM-based steam turbine sub-network consists of 1 sequence input layer, 30 LSTM layers, 3 full connection layers, 1 Drop layer and 1 sequence output layer. The structure of the LSTM units in the LSTM-based steam turbine sub-network is shown in FIG. 2. The calculation method of the standard deviation normalization is the same as the calculation method in step S2.
S9, combining the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network to obtain a complete primary frequency regulation load prediction network of the deep peaking thermal power generating unit;
the structure of the complete depth peaking thermal power generating unit primary frequency modulation load prediction network is shown in fig. 3.
S10, inputting the operation historical data of the deep peaker adjusting thermal power generating unit into a deep peaker adjusting thermal power generating unit primary frequency modulation load prediction network, and realizing short-term prediction of future main steam pressure, main steam temperature, reheat steam pressure, reheat steam temperature, regulation stage pressure and unit load;
And S11, after the unit acquires new sampling data, adding the new sampling data into the input data, removing data with the same time length at the initial moment from the original input data, updating the network state, and repeating the S10 process to realize the on-line prediction of the primary frequency modulation load of the deep peaking thermal power unit.
In another embodiment of the present invention, an LSTM-based online estimation system for primary frequency modulation capability of a depth peak shaver set is provided, which can be used to implement the LSTM-based online estimation method for primary frequency modulation capability of a depth peak shaver set.
The network training module respectively constructs a boiler combustion system sub-network based on an LSTM neural network, a reheating system sub-network based on the LSTM neural network, a steam turbine speed regulation system sub-network and a steam turbine sub-network based on the LSTM neural network, constructs a time sequence input sequence and a time sequence target sequence with uniform time dimension according to the operation historical data of the deep peaking thermal power generating unit, and respectively trains the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network by utilizing the corresponding time sequence input sequence and the time sequence target sequence;
The network combination module is used for combining the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network which are trained by the network training module to obtain a complete primary frequency modulation load prediction network of the deep peaking thermal power generating unit;
the short-term prediction module inputs the operation historical data of the deep peaking thermal power generating unit into a primary frequency modulation load prediction network of the deep peaking thermal power generating unit obtained by the network combination module, and performs short-term prediction on the main steam pressure, the main steam temperature, the reheated steam pressure, the reheated steam temperature, the regulating stage pressure and the unit load of a future unit;
and the online prediction module is used for adding the new sampling data into the input data after the unit acquires the new sampling data, removing the data with the same time length at the initial moment from the original input data, updating the network state, and repeating the short-term prediction module to realize the online prediction of the primary frequency modulation load of the deep peaking thermal power unit.
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 drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be obtained by a person skilled in the art without inventive step based on the embodiments of the present invention, are within the scope of protection of the present invention.
Examples
Referring to fig. 4, the present example is based on the historical data of the operation of the crane-wall-tornado crane-power generation limited model No. 2, and a simulation experiment for estimating the primary frequency modulation capability is performed.
The rated load of the unit 2 is 600MW, a model of a 9-type speed governor in a power system analysis integration program PSASP is taken as a traditional method, and compared with the method provided by the invention, the primary frequency modulation process under 33% of rated load is estimated, and the result is shown in fig. 4.
In the figure, the estimation is carried out after 32s, and the result shows that in the estimation process, the maximum error of the estimation result of the method is 0.48MW, and the maximum error of the estimation result of the traditional method is 2.38MW, and the maximum error is reduced 79.83, which shows that the method has higher accuracy.
In summary, according to the LSTM-based online estimation method and system for the primary frequency modulation capability of the deep peaking generation unit, the unique advantages of the LSTM deep neural network in processing the timing sequence problem are utilized, the primary frequency modulation characteristics of the equipment and links influencing the primary frequency modulation of the thermal power generation unit under the deep peaking working condition are captured, the online prediction of the primary frequency modulation load of the deep peaking generation unit is achieved, compared with the traditional technology, all the equipment and links influencing the primary frequency modulation of the thermal power generation unit are considered, information contained in historical operation data of the thermal power generation unit is fully mined, and the prediction accuracy is remarkably improved.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above contents are only for illustrating the technical idea of the present invention, and the protection scope of the present invention should not be limited thereby, and any modification made on the basis of the technical idea proposed by the present invention falls within the protection scope of the claims of the present invention.

Claims (10)

1. The LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method is characterized by comprising the following steps of:
Respectively constructing a boiler combustion system sub-network based on an LSTM neural network, a reheating system sub-network based on the LSTM neural network, a steam turbine speed regulation system sub-network based on an MLP (maximum likelihood prediction) and a steam turbine sub-network based on the LSTM neural network, constructing a time sequence input sequence and a time sequence target sequence with uniform time dimension according to the operation historical data of the deep peaking thermal power generating unit, and respectively training the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network by utilizing the corresponding time sequence input sequence and the time sequence target sequence;
combining the trained boiler combustion system sub-network, reheating system sub-network, steam turbine speed regulation system sub-network and steam turbine sub-network to obtain a complete primary frequency regulation load prediction network of the deep peaking thermal power generating unit;
inputting the operation historical data of the deep peaking generation unit into the obtained primary frequency modulation load prediction network of the deep peaking generation unit, and performing short-term prediction on main steam pressure, main steam temperature, pressure after valve adjustment, reheat steam pressure, reheat steam temperature and unit load of a future unit;
and after the deep peaking thermal power generating unit acquires new sampling data, adding the new sampling data into the input data, removing data with the same time length at the initial moment from the original input data, updating the network state, repeating short-term prediction, and realizing the on-line prediction of the primary frequency modulation load of the deep peaking thermal power generating unit.
2. The LSTM-based depth peaking unit primary frequency modulation capability online estimation method according to claim 1, wherein the boiler combustion system sub-network based on LSTM neural network is constructed by the following specific steps:
constructing a fuel quantity, a valve regulating instruction, a water supply quantity and a air supply quantity in the operation historical data of the deep peaking thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing main steam pressure and main steam temperature in the operation historical data of the deep peaking thermal power generating unit into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the time sequence target sequence have the same time length, the time sequence target sequence lags behind the time sequence input sequence, and the lag time length is the time length to be predicted;
constructing a boiler combustion system sub-network based on an LSTM neural network, standardizing the constructed time sequence input sequence and the time sequence target sequence by adopting a standard deviation method, and training the boiler combustion system sub-network based on the LSTM by using the standardized time sequence input sequence and the standardized time sequence target sequence to obtain the trained boiler combustion system sub-network.
3. The LSTM-based depth peaking unit primary frequency modulation capacity online estimation method according to claim 1, wherein the construction of the LSTM neural network-based reheating system sub-network is specifically as follows:
Constructing the fuel quantity, the valve adjusting instruction, the water supply quantity and the air supply quantity in the operation historical data of the deep peaking thermal power generating unit into a time sequence input sequence with a uniform time dimension; constructing a reheat steam pressure and a reheat steam temperature in the operation historical data of the deep peaking thermal power generating unit into a time sequence target sequence with a uniform time dimension; the time sequence input sequence and the time sequence target sequence have the same time length, the time sequence target sequence lags behind the time sequence input sequence, and the lag time length is the time length to be predicted;
constructing a reheating system sub-network based on an LSTM neural network, standardizing the constructed time sequence input sequence and the time sequence target sequence by adopting a standard deviation method, and training the reheating system sub-network based on the LSTM by using the standardized time sequence input sequence and the standardized time sequence target sequence to obtain the trained reheating system sub-network.
4. The LSTM-based online estimation method for the primary frequency modulation capacity of the depth peak shaver set according to claim 1, wherein the step of constructing a steam turbine speed regulation system sub-network specifically comprises the following steps:
constructing a valve regulating instruction, main steam pressure and main steam temperature in operation historical data of a deep peaker thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing regulating stage pressure in the operation historical data of the deep peaker thermal power generating unit into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the time sequence target sequence have the same time length, and the time of the time sequence input sequence corresponds to the time of the time sequence target sequence one by one;
And constructing a steam turbine speed regulating system sub-network, standardizing the constructed time sequence input sequence and the time sequence target sequence by adopting a standard deviation method, and training the steam turbine speed regulating system sub-network by using the standardized time sequence input sequence and the standardized time sequence target sequence to obtain the trained steam turbine speed regulating system sub-network.
5. The LSTM-based online estimation method for primary frequency modulation capability of depth peak shaving unit according to claim 4, wherein the turbine speed regulation system is a subsystemThe network comprises a sequence input layer, a plurality of full connection layers, a plurality of Drop layers and a sequence output layer, wherein the full connection layer Yt＝WXt+ b, W represents the fully connected layer learnable weight, b represents the learnable bias weight, XtIndicating the input of the fully connected layer at time t, YtRepresenting the output of the fully connected layer at time t.
6. The LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method according to claim 1, wherein the LSTM neural network-based steam turbine sub-network is constructed by the following specific steps:
constructing a regulating stage pressure, a reheat steam pressure and a reheat steam temperature in the operation historical data of the deep peaking thermal power generating unit into a time sequence input sequence with a uniform time dimension, constructing a unit load in the operation historical data of the deep peaking thermal power generating unit into a time sequence target sequence with the uniform time dimension, wherein the time sequence input sequence and the time sequence target sequence have the same time length, and the time of the time sequence input sequence corresponds to the time of the time sequence target sequence one by one;
And constructing a steam turbine sub-network based on the LSTM neural network, standardizing the constructed time sequence input sequence and the target sequence by adopting a standard deviation method, and training the steam turbine sub-network based on the LSTM by using the standardized time sequence input sequence and the standardized time sequence target sequence to obtain the trained steam turbine sub-network.
7. The LSTM-based depth peaking unit primary frequency modulation capability online estimation method according to any one of claims 1 to 6, wherein the time sequence input sequence with uniform time dimension is:
Figure FDA0003654746690000031
the timing target sequence with uniform time dimension is:
Figure FDA0003654746690000032
N1、N2respectively representing the data types in input sequence and target sequenceAnd the time sequence input sequence comprises original data values with T time lengths and data change values with T time lengths.
8. The LSTM-based depth peaking unit chirp capacity online estimation method of claim 1, wherein the LSTM neural network-based boiler combustion system sub-network, the LSTM neural network-based reheat system sub-network and the LSTM neural network-based steam turbine sub-network comprise a sequence input layer, a plurality of LSTM layers, a plurality of full connection layers, a plurality of Drop layers and a sequence output layer respectively.
9. The LSTM-based online estimation method for the primary frequency modulation capability of the depth peak shaver set as claimed in claim 8, wherein the LSTM memory unit of the LSTM layer is calculated as follows:
Figure FDA0003654746690000041
ct＝ft⊙ct-1+it⊙gt
ht＝ot⊙σc(ct)
it＝σg(Wixt+Riht-1+bi)
ft＝σg(Wfxt+Rfht-1+bf)
gt＝σc(Wgxt+Rght-1+bg)
ot＝σg(Woxt+Roht-1+bo)
σ(x)＝(1+e-x)-1
wherein W represents a learnable input weight, R represents a learnable cyclic weight, b represents a learnable bias; i. f, g and o respectively represent an input gate,A forgetting gate, a candidate unit and an output gate; c. CtRepresents the cell state at time step t; h istRepresenting the hidden state at time step t; sigmacRepresenting the tan h state activation function; sigmagRepresenting a sigmoid gate activation function; full connection layer Yt＝WXt+ b, W represents the fully connected layer learnable weight, b represents the learnable bias weight, XtIndicating the input of the fully-connected layer at time t, YtRepresenting the output of the fully connected layer at time t.
10. An LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation system is characterized by comprising:
the network training module is used for respectively constructing a boiler combustion system sub-network based on an LSTM neural network, a reheating system sub-network based on the LSTM neural network, a steam turbine speed regulation system sub-network and a steam turbine sub-network based on the LSTM neural network, constructing a time sequence input sequence and a time sequence target sequence with uniform time dimension according to the operation historical data of the deep peak regulation thermal power generating unit, and respectively training the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network by utilizing the corresponding time sequence input sequence and the time sequence target sequence;
The network combination module is used for combining the boiler combustion system sub-network, the reheating system sub-network, the steam turbine speed regulation system sub-network and the steam turbine sub-network which are trained by the network training module to obtain a complete primary frequency modulation load prediction network of the deep peaking thermal power generating unit;
the short-term prediction module is used for inputting the operation historical data of the deep peak shaving thermal power generating unit into a primary frequency modulation load prediction network of the deep peak shaving thermal power generating unit obtained by the network combination module, and performing short-term prediction on the main steam pressure, the main steam temperature, the reheated steam pressure, the reheated steam temperature, the regulation stage pressure and the unit load of a future unit;
and the online prediction module is used for adding the new sampling data into the input data after the unit acquires the new sampling data, eliminating the data with the same time length at the initial moment from the original input data, updating the network state, and repeating the short-term prediction module to realize the online prediction of the primary frequency modulation load of the deep peak-shaving thermal power generating unit.
CN202210550300.1A 2022-05-20 2022-05-20 LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system Pending CN114759613A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210550300.1A CN114759613A (en) 2022-05-20 2022-05-20 LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210550300.1A CN114759613A (en) 2022-05-20 2022-05-20 LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system

Publications (1)

Publication Number Publication Date
CN114759613A true CN114759613A (en) 2022-07-15

Family

ID=82334967

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210550300.1A Pending CN114759613A (en) 2022-05-20 2022-05-20 LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system

Country Status (1)

Country Link
CN (1) CN114759613A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117869930A (en) * 2024-01-31 2024-04-12 中国电力工程顾问集团有限公司 Multi-variable control method and device for stable combustion of coal-fired boiler in wide load range

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117869930A (en) * 2024-01-31 2024-04-12 中国电力工程顾问集团有限公司 Multi-variable control method and device for stable combustion of coal-fired boiler in wide load range
CN117869930B (en) * 2024-01-31 2024-06-07 中国电力工程顾问集团有限公司 Multi-variable control method and device for stable combustion of coal-fired boiler in wide load range

Similar Documents

Publication Publication Date Title
CN110285403A (en) Main Steam Temperature Control method based on controlled parameter prediction
CN110262248B (en) Fault robust self-adaptive reconstruction method for micro gas turbine
CN109066651B (en) Method for calculating limit transmission power of wind power-load scene
Wen et al. Data‐driven transient frequency stability assessment: A deep learning method with combined estimation‐correction framework
CN114759613A (en) LSTM-based depth peak shaving unit primary frequency modulation capacity online estimation method and system
Cassamo et al. Model predictive control for wake redirection in wind farms: a koopman dynamic mode decomposition approach
CN111931436A (en) Burner nozzle air quantity prediction method based on numerical simulation and neural network
Yao et al. Fault diagnosis based on RseNet-LSTM for industrial process
CN111680823A (en) Wind direction information prediction method and system
Farag et al. Design and implementation of a variable-structure adaptive fuzzy-logic yaw controller for large wind turbines
CN113505525A (en) Power system dynamic element modeling method and device based on differential neural network
CN107527093B (en) Wind turbine generator running state diagnosis method and device
Carrasco et al. Novel modelling for a steam boiler under fast load dynamics with implications to control
CN117394313A (en) Power system transient stability evaluation method, system, chip and equipment
Ma et al. Prediction of thermal system parameters based on PSO-ELM hybrid algorithm
CN116522752A (en) Compressed air energy storage system simulation method based on mechanism and data fusion
CN113111588B (en) NO of gas turbine X Emission concentration prediction method and device
Hua et al. Distributed control for uncertain nonlinear multiagent systems subject to hybrid faults
CN114996863A (en) Turbofan engine T-S fuzzy modeling method based on feature extraction
Ma et al. Inverse control for the coordination system of supercritical power unit based on dynamic fuzzy neural network modeling
CN110631003B (en) Reheated steam temperature adjusting method based on hierarchical scheduling multi-model predictive control
Grelewicz et al. Practical Verification of the Advanced Control Algorithms Based on the Virtual Commissioning Methodology-A Case Study
Jiang et al. Application of an optimized grey system model on 5-Axis CNC machine tool thermal error modeling
Kalabić et al. Decentralized constraint enforcement using reference governors
Lei et al. Research on Intelligent PID Control Algorithm Based on Neural Network

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