CN114896891A - Steam simulation calculation method based on error correction of nuclear extreme learning machine - Google Patents
Steam simulation calculation method based on error correction of nuclear extreme learning machine Download PDFInfo
- Publication number
- CN114896891A CN114896891A CN202210600076.2A CN202210600076A CN114896891A CN 114896891 A CN114896891 A CN 114896891A CN 202210600076 A CN202210600076 A CN 202210600076A CN 114896891 A CN114896891 A CN 114896891A
- Authority
- CN
- China
- Prior art keywords
- steam
- parameters
- learning machine
- error
- calculation
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Geometry (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Hardware Design (AREA)
- Fluid Mechanics (AREA)
- Physiology (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Algebra (AREA)
- Mathematical Analysis (AREA)
- Genetics & Genomics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a steam simulation calculation method based on error correction of a nuclear limit learning machine, aiming at the problem of poor model construction accuracy in the conventional steam heating network simulation method. Firstly, acquiring pipeline parameters and steam parameters are confirmed, the parameters are divided into training samples for establishing a steam-water thermal coupling simulation model for calculation, errors obtained by verifying simulation results by using real pipe network data are used as output samples of a genetic algorithm-kernel limit learning machine error prediction model for training. And when the steam heat supply network model operates, calling the trained error prediction model to perform error prediction, so that the steam pipe network water-heat power coupling simulation calculation result is corrected, and the precision of the steam heat supply network simulation model is improved. The method is suitable for dynamic simulation of the steam heating network, and can be used for optimizing an original steam heating network simulation model in a targeted manner by utilizing a large amount of accumulated pipe network operation data of a heating enterprise all year round and improving the precision of a simulation result.
Description
Technical Field
The invention relates to the technical field of energy system simulation, in particular to a steam simulation calculation method based on error correction of a nuclear limit learning machine.
Background knowledge
With the continuous optimized land layout of cities, the cluster-type industrial park becomes an important way for pulling the regional economy of China to grow rapidly and pushing the regional technical innovation. However, since the construction of the industrial park enterprises is not a uniform planning implementation, but is formed by not calculating the industrial expansion and the structural reorganization according to the requirements, the non-systematic construction of the buildings brings great challenges to the dynamic construction of the heating system of the industrial park. Under the background, how to solve the problems of local pipeline overload, supply and demand mismatching and the like under the conditions that the incremental capacity of a heating system and the structure of a heat supply network are increasingly complicated by using a newly-built pipe network intelligent system becomes a difficult problem which needs to be overcome by the current comprehensive energy system.
Aiming at the outstanding safety and quality guarantee problems of the complex steam heat supply pipe network in the industrial park, the traditional heat supply system mostly adjusts the flow conveying of the pipe network by depending on experience, and the energy waste phenomenon is serious. Therefore, an intelligent heat supply concept is deeply applied to an industrial park taking central heat supply as a main heat supply mode, namely, heat supply planning based on a model is taken as a technical core, machine calculation based on an intelligent optimization algorithm is taken as a technical support, and the intelligent and fine development of heat supply system planning is promoted. In recent years, a central Heating technology (GDH) is gradually developed, and GDH is facing the challenge of making buildings more energy-saving and becomes a component of the operation of intelligent energy systems, i.e. an integrated intelligent power, gas and heat supply network. Due to the progress of the electronic communication technology and the more mature technology of the internet of things, big data and cloud computing, powerful software and hardware support is provided for the improvement of the operation regulation level of the central heating system. The improvement of the operation regulation level of the central heating system can not leave the research of theories and methods such as the mathematical modeling, the dynamic characteristic analysis and the operation control of the central heating system.
The scholars at home and abroad provide various researches with the attention on the construction model, model optimization and the like of the regional steam heating system. Dividing the nodes of the pipe network from three conservation equations (supplementing a state equation if necessary), dispersing the conservation equations on the nodes, and then solving an equation set to obtain the pressure and the temperature of the nodes; or establishing a transient model capable of reflecting the fluctuation of the running state of the steam pipe network on the basis of a transient flow theory and a characteristic line solving method; in another dynamic simulation method for a steam heating network based on internal conservation, a dynamic simulation mathematical model of the hot water heating network is established, partial differential equations of dynamic variation characteristics of pipeline pressure are differentiated into an algebraic equation set, and state information of the whole pipe network and the like are iteratively obtained and output through the relationship of internal conservation.
The simulation methods described above are all calculated according to the mechanism of steam flowing in the pipeline, but due to the existence of factors such as pipe wall scaling and construction installation, the steam operation parameters of the actual pipeline always have certain deviation from the mechanism calculation result. The calculation precision of the steam pipeline can be effectively improved by reducing the error of the part, and a better reference is provided for the operation optimization of the heat supply pipe network.
Disclosure of Invention
The invention provides a steam simulation calculation method based on error correction of a nuclear limit learning machine, aiming at the problem of poor model construction accuracy in the conventional steam heating network simulation method. Firstly, acquiring pipeline parameters and steam parameters are confirmed, the parameters are divided into training samples for establishing a steam-water thermal coupling simulation model for calculation, errors obtained by verifying simulation results by using real pipe network data are used as output samples of a genetic algorithm-kernel limit learning machine error prediction model for training. And when the steam heat network model runs, calling the trained error prediction model to perform error prediction. Therefore, the steam pipe network water-heat power coupling simulation calculation result is corrected, and the precision of the steam heat network simulation model is improved. The method is suitable for dynamic simulation of the steam heating network, and can be used for optimizing an original steam heating network simulation model in a targeted manner by utilizing a large amount of accumulated pipe network operation data of a heating enterprise all year round and improving the precision of a simulation result.
In order to achieve the purpose, the invention adopts the technical scheme that: the steam simulation calculation method based on the error correction of the nuclear limit learning machine comprises the following steps:
and S1, confirming the parameters of the collecting pipeline and the steam parameters, and randomly dividing the parameters into training samples and testing samples. And establishing a hydraulic thermal coupling model for the training sample for calculation, and subtracting the calculation result from the actual measurement value to obtain the mechanism error of the training sample.
And S2, training the error prediction model of the genetic algorithm-kernel limit learning machine by taking the pipeline parameters and the steam parameters of the training samples as inputs and taking the mechanism error as an output.
And S3, the test sample is used for hydraulic and thermal coupling calculation, and a mechanism calculation result is obtained through calculation. And predicting the mechanism error of the test sample by the trained genetic algorithm-core extreme learning machine model. And correcting the error predicted value of the test sample by the hydraulic-thermal coupling calculation result of the test sample to obtain the output of the mixed model.
As a calculation method of the present invention, the step S1 of confirming the collecting pipe parameter and the steam parameter includes:
1) the pipe diameter, the length and the topological structure of each pipeline of the pipe network. The influence of the pipe diameter on the calculation result of the steam pipe network water thermal coupling calculation is undoubted, and the flow parameter Q and the flow rate parameter V are closely related to the pipeline interface area M according to Q ═ V · M, and the area M is in a quadratic relation with the pipe diameter of the pipeline, so the pipe diameter size has great influence on the final steam pipe water thermal coupling calculation result. In addition, the length of the pipe is related to the loss of the steam pipe network along the way. The nodes and the branches form the minimum connecting element of the fluid network, a large fluid network is formed by splicing a plurality of connecting elements, and the transmission and iteration of working medium flow state parameters among different connecting elements are the basic calculation process of the steam heat supply network.
2) The pressure at the beginning and end of the steam pipe network. The pressures of the initial end and the tail end of the steam heat supply network are used as boundary condition input of the model, and the distribution of parameters such as the temperature, the flow and the pressure of the whole network steam can be obtained after iterative calculation and solution so as to be used for calculating the mechanism error of the training sample.
As a calculation method of the present invention, in step S1, a hydraulic thermal coupling model is established for a training sample to perform calculation, where the construction method of the hydraulic thermal coupling model is as follows:
the three equations are respectively a mass conservation equation, a momentum equation and an energy equation of the steam pipe network.
In the formula: ρ represents a fluid; t is time; n represents the total number of nodes; g ij Representing the mass flow between nodes i, j; v i Represents the volume of node i (volume of adjacent pipe); h ij Representing macroscopic kinetic energy and potential energy between the nodes i and j and pressure generated between power sources; u shape ij Representing the fluid flow rate between nodes i, j;
expressing the on-way resistance and the local resistance loss, wherein lambda expresses the on-way resistance coefficient, L is the length of the pipeline, d expresses the diameter of the pipeline, and epsilon expresses the equivalent resistance coefficient of the pipeline; u. of i Representing the internal energy of the node i; h is j Representing the enthalpy of node j. Q i Representing the heat of node i.
D ij Representing the connection mode between the nodes i and j, the following definitions are provided:
as a calculation method of the present invention, in step S1, the calculation result is subtracted from the actual measurement value to obtain a mechanism error of the training sample, where the method includes:
the kernel limit learning machine is a single-output prediction algorithm, so that two prediction models of flow errors and temperature errors are respectively established for predicting mechanism errors. The mechanism errors are respectively as follows: theta T ＝y′ T -y T ，θ Q ＝y′ Q -y Q . Wherein theta is T 、θ Q For errors of temperature and flow mechanism，y′ T 、y′ Q Simulation value y of temperature and flow calculated for steam pipe network water thermal power coupling T 、y Q The actual temperature value and the flow value are obtained.
As an optimization method of the present invention, in step S2, the training of the genetic algorithm-kernel limit learning machine error prediction model with the pipeline parameters and the steam parameters of the training samples as inputs and the mechanism error as an output includes:
and (3) taking the pipeline parameters and the steam parameters of the training samples as inputs and the mechanism errors as outputs to train the genetic algorithm-kernel extreme learning machine error prediction model. Because the performance of the kernel-limit learning machine is greatly influenced by the kernel function parameter gamma and the regularization coefficient C, the two parameters of the kernel-limit learning machine are optimized by adopting a genetic algorithm.
As a calculation method of the present invention, in step S3, the test sample is used for hydraulic thermal coupling calculation, and the calculated mechanism calculation result is:
taking steam parameters of a test sample as input, and solving after iterative calculation of a hydraulic coupling model to obtain a calculation result y of parameter mechanism such as temperature, flow and pressure of the steam in the whole network T y Q Distribution of (2).
As a calculation method of the present invention, the step S3 of predicting the mechanism error of the trained genetic algorithm-kernel-limit learning machine model on the test sample includes:
taking the pipeline parameters and steam parameters of a test sample as input, and calculating a mechanism error predicted value theta of the sample through a genetic algorithm-kernel limit learning machine error prediction model T θ Q 。
As a calculation method of the present invention, the error prediction value of the test sample in the step S3 corrects the calculation result of the hydraulic thermal coupling of the test sample:
correcting the error predicted value of the test sample to obtain a hydraulic thermal coupling calculation result y finl ＝y+θ。
Drawings
FIG. 1 is a flow chart of the steps of the method of the present invention;
FIG. 2 is a schematic view of the topology of the pipe network structure of the present invention;
FIG. 3 is a flow chart of a steam pipe network water thermal power coupling calculation model of the present invention;
FIG. 4 is a flow chart of the genetic algorithm optimized kernel-limit learning machine of the present invention;
fig. 5 is a calculation result of the embodiment of the present invention.
Detailed Description
The invention will be further explained by the embodiments in the following with reference to the drawings. However, this embodiment is merely illustrative, and the scope of the present invention is not limited by this embodiment.
A flow chart of the steam simulation calculation method based on error correction of the nuclear limit learning machine is shown in FIG. 1, and the method comprises the following steps:
s1, confirming that the collecting pipeline parameters and the steam parameters comprise:
1) the pipe diameter, the length and the topological structure of each pipeline of the pipe network. The influence of the pipe diameter on the calculation result of the steam pipe network water thermal coupling calculation is undoubted, and the flow parameter Q and the flow rate parameter V are closely related to the pipeline interface area M according to Q ═ V · M, and the area M is in a quadratic relation with the pipe diameter of the pipeline, so the pipe diameter size has great influence on the final steam pipe water thermal coupling calculation result. In addition, the pipe length is related to the loss of the steam pipe network along the way. The nodes and the branches form the minimum connecting element of the fluid network, a large fluid network is formed by splicing a plurality of connecting elements, the transmission and iteration of working medium flow state parameters among different connecting elements are the basic calculation process of the steam heat supply network, and a simplified steam pipe network topological structure is shown in fig. 2.
2) The pressure at the beginning and end of the steam pipe network. The pressures of the initial end and the tail end of the steam heat supply network are used as boundary condition input of the model, and the distribution of parameters such as the temperature, the flow and the pressure of the whole network steam can be obtained after iterative calculation and solution so as to be used for calculating the mechanism error of the training sample. FIG. 3 is a flow chart of a steam pipe network water thermal coupling calculation model. The specific construction method further comprises the following steps:
s1, calculating the training sample, namely the collected historical data, by adopting a hydraulic thermal coupling model, wherein the construction method of the hydraulic thermal coupling model comprises the following steps:
the three equations are respectively a mass conservation equation, a momentum equation and an energy equation of the steam pipe network.
In the formula: ρ represents a fluid; t is time; n represents the total number of nodes; g ij Representing the mass flow between nodes i, j; v i Represents the volume of node i (volume of adjacent pipe); h ij Representing macroscopic kinetic energy and potential energy between the nodes i and j and pressure generated between power sources; u shape ij Representing the fluid flow rate between nodes i, j;
expressing the on-way resistance and the local resistance loss, wherein lambda expresses the on-way resistance coefficient, L is the length of the pipeline, d expresses the diameter of the pipeline, and epsilon expresses the equivalent resistance coefficient of the pipeline; u. of i Representing the internal energy of the node i; h is j Representing the enthalpy of node j. Q i Indicating the heat of node i.
D ij Representing the connection mode between the nodes i and j, the following definitions are provided:
the three formulas are solved simultaneously. Because the three formulas are completely coupled, in the simulation process, the pressure, the flow and the enthalpy of each node can be obtained through solving, so that the temperature can be deduced. Thereby obtaining the flow and temperature distribution of each node of the pipe network.
And S1, subtracting the calculation result from the actual measurement value to obtain the mechanism error of the training sample. The kernel limit learning machine is a single-output prediction algorithm, so that two prediction models of flow errors and temperature errors are respectively established for predicting mechanism errors. The mechanism errors are respectively as follows: theta T ＝y′ T -y T ，θ Q ＝y′ Q -y Q . Wherein theta is T 、θ Q Is temperature, flow mechanism error, y' T 、y′ Q Simulation value y of temperature and flow calculated for steam pipe network water thermal power coupling T 、y Q The actual temperature value and the flow value are obtained.
And S2, training the error prediction model of the genetic algorithm-kernel limit learning machine by taking the pipeline parameters and the steam parameters of the training samples as inputs and taking the mechanism error as an output. The performance of the kernel-limit learning machine is greatly influenced by the kernel function parameter gamma and the regularization coefficient C, so that the two parameters of the kernel-limit learning machine are optimized by adopting a genetic algorithm, a flow chart of the optimization is shown in FIG. 4, and the main steps are as follows:
s21, determining a Root Mean Square Error (RMSE) of the prediction as a fitness function: and training the kernel extreme learning machine network by adopting a 10-order cross verification method, and calculating the root mean square error of the prediction result as a fitness function value. Equation of root mean square error
In the formula, y i In the form of an actual value of the value,
is a predicted value.
S22, setting genetic algorithm parameters: cross probability, mutation probability, population size, and genetic algebra.
And S23, calculating the fitness function values of all individuals in the population.
And S24, performing selection, crossing and mutation operations to form a new population.
S25, judging whether the termination condition is satisfied, if not, returning to the step S23; and if the termination condition is met, obtaining the optimized parameters. The termination condition is set to reach the maximum genetic algebra.
And S26, predicting the test sample by using the optimized parameters by using the kernel limit learning machine.
S3, using the test sample, namely the field data, for the waterpowerAnd (4) calculating the thermodynamic coupling to obtain a mechanism calculation result. Taking steam parameters of a test sample as input, and solving after iterative calculation of a hydraulic coupling model to obtain a calculation result y of parameter mechanism such as temperature, flow and pressure of the steam in the whole network T y Q Distribution of (2). The calculation step is the same as S1.
And S3, predicting the mechanism error of the test sample by using the trained genetic algorithm-kernel limit learning machine model. Taking the pipeline parameters and steam parameters of a test sample as input, and calculating a mechanism error predicted value theta of the sample through a genetic algorithm-kernel limit learning machine error prediction model T θ Q 。
S3, correcting the error prediction value of the test sample to obtain the result y of the hydraulic thermal coupling calculation of the test sample finl ＝y+θ。
The invention provides a steam simulation calculation method based on error correction of a nuclear limit learning machine, aiming at the problem of poor model construction accuracy in the conventional steam heating network simulation method. Firstly, a simulation model of a steam heat supply network is established, the hydraulic and thermal coupling calculation of the steam pipe network is carried out by taking historical data as the root, and a genetic algorithm-kernel limit learning machine error prediction model is established for a training sample according to the error of an output result and an actual result. And carrying out steam hydraulic thermal coupling calculation on the test sample when the actual steam simulation pipe network runs, and calling a genetic algorithm-kernel limit learning machine error prediction model to obtain an excitation error prediction value for correcting a calculation result, so that the aim of improving the precision of the steam heat network simulation model is fulfilled. The method is suitable for dynamic simulation of the steam heating network, and can be used for optimizing an original steam heating network simulation model in a targeted manner by utilizing a large amount of accumulated pipe network operation data of a heating enterprise all year round and improving the precision of a simulation result. The flow calculation result of the original simulation model and the simulation result corrected by the error of the kernel limit learning machine are shown in fig. 5, and the calculation precision is greatly improved.
The above embodiments are only exemplary embodiments of the present application, and are not intended to limit the present application, and the protection scope of the present application is defined by the claims. Various modifications and equivalents may be made by those skilled in the art within the spirit and scope of the present application and such modifications and equivalents should also be considered to be within the scope of the present application.
Claims (9)
1. A steam simulation calculation method based on error correction of a nuclear extreme learning machine is characterized by comprising the following steps:
and S1, confirming the parameters of the collecting pipeline and the steam parameters, and randomly dividing the parameters into training samples and testing samples. And establishing a hydraulic thermal coupling model for the training sample for calculation, and subtracting the calculation result from the actual measurement value to obtain the mechanism error of the training sample.
And S2, training the error prediction model of the genetic algorithm-kernel limit learning machine by taking the pipeline parameters and the steam parameters of the training samples as inputs and taking the mechanism error as an output.
And S3, the test sample is used for hydraulic and thermal coupling calculation, and a mechanism calculation result is obtained through calculation. And predicting the mechanism error of the test sample by the trained genetic algorithm-kernel limit learning machine model. And correcting the error predicted value of the test sample by the hydraulic-thermal coupling calculation result of the test sample to obtain the output of the mixed model.
2. The method for steam simulation calculation based on error correction of the nuclear limit learning machine according to claim 1, wherein the step S1 of confirming the collection pipeline parameters and the steam parameters comprises:
1) the pipe diameter, the length and the topological structure of each pipeline of the pipe network. The influence of the pipe diameter on the calculation result of the steam pipe network water thermal coupling calculation is undoubted, and the flow parameter Q and the flow rate parameter V are closely related to the pipeline interface area M according to Q ═ V · M, and the area M is in a quadratic relation with the pipe diameter of the pipeline, so the pipe diameter size has great influence on the final steam pipe water thermal coupling calculation result. In addition, the length of the pipe is related to the loss of the steam pipe network along the way. The nodes and the branches form the minimum connecting element of the fluid network, a large fluid network is formed by splicing a plurality of connecting elements, and the transmission and iteration of working medium flow state parameters among different connecting elements are the basic calculation process of the steam heat supply network.
2) The pressure at the beginning and end of the steam pipe network. The pressures of the initial end and the tail end of the steam heat supply network are used as boundary condition input of the model, and the distribution of parameters such as the temperature, the flow and the pressure of the whole network steam can be obtained after iterative calculation and solution so as to be used for calculating the mechanism error of the training sample.
3. The method for calculating the hydraulic thermal coupling model established by the training samples according to claim 1, wherein the method for establishing the hydraulic thermal coupling model in step S1 is as follows:
the three equations are respectively a mass conservation equation, a momentum equation and an energy equation of the steam pipe network.
In the formula: ρ represents a fluid; t is time; n represents the total number of nodes; g ij Representing the mass flow between nodes i, j; v i Represents the volume of node i (volume of adjacent pipe); h ij Representing macroscopic kinetic energy and potential energy between the nodes i and j and pressure generated between power sources; u shape ij Representing the fluid flow rate between nodes i, j;
expressing the on-way resistance and the local resistance loss, wherein lambda expresses the on-way resistance coefficient, L is the length of the pipeline, d expresses the diameter of the pipeline, and epsilon expresses the equivalent resistance coefficient of the pipeline; u. of i Representing the internal energy of the node i; h is j Representing the enthalpy of node j. Q i Representing the heat of node i.
D ij Representing the connection mode between the nodes i and j, the following definitions are provided:
the three formulas are solved simultaneously. Because the three formulas are completely coupled, in the simulation process, the pressure, the flow and the enthalpy of each node can be obtained through solving, so that the temperature can be deduced. Thereby obtaining the flow and temperature distribution of each node of the pipe network.
4. The steam simulation calculation method based on error correction of the kernel-limit learning machine as claimed in claim 1, wherein the calculated result is subtracted from the actual measured value in step S1 to obtain the mechanism error of the training sample:
because the kernel limit learning machine is a single-output prediction algorithm, two prediction models of flow errors and temperature errors are respectively established for predicting mechanism errors. The mechanism errors are respectively as follows: theta T ＝y′ T -y T ，θ Q ＝y′ Q -y Q . Wherein theta is T 、θ Q Is temperature, flow mechanism error, y' T 、y′ O Simulation value y of temperature and flow calculated for steam pipe network water thermal power coupling T 、y Q The actual temperature value and the flow value are obtained.
5. The steam simulation calculation method based on error correction of the nuclear limit learning machine according to claims 1 and 4, characterized in that the training of the genetic algorithm-nuclear limit learning machine error prediction model is performed by taking the pipeline parameters and the steam parameters of the training samples as inputs and the mechanism error as an output in step S2 of claim 1:
and (3) taking the pipeline parameters and the steam parameters of the training samples as inputs and the mechanism errors as outputs to train the genetic algorithm-kernel extreme learning machine error prediction model. Because the performance of the kernel-limit learning machine is greatly influenced by the kernel function parameter gamma and the regularization coefficient C, the two parameters of the kernel-limit learning machine are optimized by adopting a genetic algorithm.
6. The method for steam simulation calculation based on error correction of the nuclear learning machine according to claim 1 or 5, wherein the training of the genetic algorithm-the nuclear learning machine error prediction model in the step S2 further comprises:
s21, determining a Root Mean Square Error (RMSE) of the prediction as a fitness function: and training the kernel extreme learning machine network by adopting a 10-order cross verification method, and calculating the root mean square error of the prediction result as a fitness function value. Equation of root mean square error
In the formula, y i In the form of an actual value of the value,
is a predicted value.
S22, setting genetic algorithm parameters: cross probability, mutation probability, population size, and genetic algebra.
And S23, calculating the fitness function values of all individuals in the population.
And S24, performing selection, crossing and mutation operations to form a new population.
S25, judging whether the termination condition is satisfied, if not, returning to the step S23; and if the termination condition is met, obtaining the optimized parameters. The termination condition is set to reach the maximum genetic algebra.
And S26, predicting the test sample by using the optimized parameters by using the kernel limit learning machine.
7. The method for steam simulation calculation based on error correction of the nuclear limit learning machine according to claim 1, wherein the step S3 uses the test sample, i.e. the field data, for the hydrothermodynamic coupling calculation to obtain the mechanism calculation result, and the step S3 includes:
taking steam parameters of a test sample as input, and solving after iterative calculation of a hydraulic coupling model to obtain a calculation result y of parameter mechanism such as temperature, flow and pressure of the steam in the whole network T y Q Distribution of (2).
8. The steam simulation calculation method based on error correction of the nuclear learning machine according to claim 1, wherein the step S3 of predicting the mechanism error of the trained genetic algorithm-nuclear learning machine model on the test sample comprises:
taking a test sample, namely pipeline parameters and steam parameters of field data as input, and calculating a mechanism error predicted value theta of the sample through a genetic algorithm-kernel limit learning machine error prediction model T θ Q 。
9. The method for steam simulation calculation based on error correction of the nuclear limit learning machine according to claim 1, wherein the step S3 of correcting the result of the calculation of the hydraulic thermal coupling of the test sample by the predicted error value of the test sample comprises:
correcting the error predicted value of the test sample to obtain a hydraulic thermal coupling calculation result y finl ＝y+θ。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210600076.2A CN114896891A (en) | 2022-05-27 | 2022-05-27 | Steam simulation calculation method based on error correction of nuclear extreme learning machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210600076.2A CN114896891A (en) | 2022-05-27 | 2022-05-27 | Steam simulation calculation method based on error correction of nuclear extreme learning machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114896891A true CN114896891A (en) | 2022-08-12 |
Family
ID=82725769
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210600076.2A Pending CN114896891A (en) | 2022-05-27 | 2022-05-27 | Steam simulation calculation method based on error correction of nuclear extreme learning machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114896891A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117010131A (en) * | 2023-09-14 | 2023-11-07 | 华能苏州热电有限责任公司 | Method for analyzing network management loss of steam heating pipe |
| CN117272560A (en) * | 2023-09-15 | 2023-12-22 | 华能苏州热电有限责任公司 | Quick building method and device for heating steam pipe network simulation model |
-
2022
- 2022-05-27 CN CN202210600076.2A patent/CN114896891A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117010131A (en) * | 2023-09-14 | 2023-11-07 | 华能苏州热电有限责任公司 | Method for analyzing network management loss of steam heating pipe |
| CN117272560A (en) * | 2023-09-15 | 2023-12-22 | 华能苏州热电有限责任公司 | Quick building method and device for heating steam pipe network simulation model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111353652B (en) | Wind power output short-term interval prediction method | |
| Zhou et al. | Dynamic simulation of natural gas pipeline network based on interpretable machine learning model | |
| CN110414114B (en) | Multi-objective multi-parameter optimization design method for U-shaped ground heat exchanger | |
| CN114896891A (en) | Steam simulation calculation method based on error correction of nuclear extreme learning machine | |
| CN110795845B (en) | Heat supply system margin quantitative calculation method and system | |
| CN112290536B (en) | Online scheduling method of electricity-heat comprehensive energy system based on near-end strategy optimization | |
| CN111046567B (en) | Godunov format-based urban drainage pipe network water flow numerical simulation method | |
| CN114548680B (en) | Automatic calibration method and system for urban storm flood management model parameters | |
| CN111783309A (en) | Dynamic simulation method of steam heating network based on internal conservation | |
| CN109992846A (en) | The emulation mode of solar cross-season underground pipe accumulation of heat | |
| CN113919107A (en) | Power grid reliability evaluation method based on maximum entropy principle and cross entropy important sampling | |
| CN114117819B (en) | Steady-state simulation method for hot steam network | |
| CN112733443A (en) | Water supply network model parameter optimization checking method based on virtual monitoring points | |
| CN114386188B (en) | Nuclear electric steam safety valve clack optimization design method based on Kerling proxy model | |
| CN113536591B (en) | Variable step dynamic simulation method for comprehensive energy system | |
| CN113361214A (en) | Open channel control model parameter identification method based on water level flow data | |
| CN103617563B (en) | A kind of water supply network theoretical based on geo-statistic spatial analysis is without monitoring node pressure determination statement | |
| Larsen et al. | Equivalent models for district heating systems | |
| CN111626003A (en) | Heating system heat load layered prediction method, storage medium and prediction equipment | |
| CN111798111A (en) | Comprehensive energy system energy supply reliability assessment method and computer system | |
| CN112113146B (en) | Synchronous self-adaptive check method for roughness coefficient and node water demand of water supply pipe network pipeline | |
| CN117371354A (en) | Natural gas pipe network transient steady-state simulation method, device, equipment and medium | |
| CN111563332B (en) | Urban water supply pipe network leakage positioning method based on night pressure regulation actual measurement | |
| CN112182905B (en) | Heat supply pipe network simulation method and device for comprehensive energy system | |
| CN113849943A (en) | Water supply network node water demand amount checking method coupled with pressure prior information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |