CN111275320B - Performance adjustment data processing method, system and storage medium of generator set - Google Patents
Performance adjustment data processing method, system and storage medium of generator set Download PDFInfo
- Publication number
- CN111275320B CN111275320B CN202010057414.3A CN202010057414A CN111275320B CN 111275320 B CN111275320 B CN 111275320B CN 202010057414 A CN202010057414 A CN 202010057414A CN 111275320 B CN111275320 B CN 111275320B
- Authority
- CN
- China
- Prior art keywords
- generator set
- data
- neural network
- operation condition
- performance
- 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.)
- Active
Links
Classifications
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- 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
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
The application discloses a method, a system and a storage medium for processing performance adjustment data of a generator set, wherein the method comprises the following steps: collecting operation condition data of each part of the generator set, and calculating performance parameters of each part of the generator set; constructing a BP neural network model according to the collected operation condition data; constructing a physical model based on the BP neural network model; predicting the operation condition of each part of the generator set through the physical model; and generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter. The application can automatically adjust the working conditions of all parts of the generator set, and simultaneously, a worker can also quickly determine the positions of devices to be adjusted according to an adjusting scheme, thereby saving adjusting and checking time and improving the working efficiency of the combined cycle generator set. The application can be widely applied to the electrical field.
Description
Technical Field
The application relates to the electrical field, in particular to a performance adjustment data processing method, a system and a storage medium of a generator set.
Background
In the using process of the gas-steam combined cycle generator set, the gas-steam combined cycle generator set is easily influenced by the external environment and the running environment inside the set, so that the gas compressor and the gas turbine are deteriorated, the running performance of the system is further influenced, and the phenomenon of mismatching of supply and demand occurs. External factors that affect the deterioration of the parts include bad air quality, gums floating in the air, vegetable gums, pollen, etc.; internal factors that affect component degradation include lubrication oil leakage, filter seal failure, component deformation, and the like. As the operating time of the unit increases, the components are subject to varying degrees of degradation, particularly in the top-cycle compressor and gas turbine components.
In a conventional combined cycle plant, in which a gas turbine is operated in a specified power state, each component is continuously degraded as the operation time of the gas turbine increases, in order to meet the requirement of the specified power even in the case of degradation, the IGV opening is usually manually adjusted, and when the requirement cannot be met by adjusting the IGV opening, the system is examined item by item from other aspects. Obviously, the traditional adjusting mode consumes more time and labor and affects the normal operation of the combined cycle generator set.
Disclosure of Invention
In order to solve the technical problems, the application aims to: provided are a performance adjustment data processing method, system and storage medium for a generator set, which can save adjustment and investigation time and improve the working efficiency of the combined cycle generator set.
A first aspect of an embodiment of the present application provides:
a performance adjustment data processing method of a generator set comprises the following steps:
collecting operation condition data of each part of the generator set, and calculating performance parameters of each part of the generator set;
constructing a BP neural network model according to the collected operation condition data;
constructing a physical model based on the BP neural network model;
predicting the operation condition of each part of the generator set through the physical model;
and generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter.
Further, the building of the BP neural network model according to the collected operation condition data specifically includes:
storing the collected operation condition data into a subset constructed in a preset time period;
acquiring operation condition data meeting preset requirements from each subset as target data;
randomly dividing the target data into training set data and test set data;
and training the neural network according to the training set data, and constructing a BP neural network model.
Further, the BP neural network model is also corrected by:
acquiring test set data in the target data;
inputting the input variable of the test set data into the BP neural network model to obtain a predicted value corresponding to the input variable of the test set data;
calculating the difference value of the predicted value corresponding to the input variable and the output variable corresponding to the input variable;
and judging whether the difference value is within an engineering acceptance range, if so, judging that the reliability of the BP neural network model passes verification, otherwise, modifying training set data and re-modeling.
Further, the building a physical model based on the BP neural network model specifically includes:
constructing a physical model according to the BP neural network model constructed in the first preset time period;
predicting the performance degradation degree of each part of the generator set in each residual preset time period through a BP neural network model constructed in the residual preset time period;
and correcting the physical model according to the performance degradation degree of each part of the generator set in each residual preset time period, and obtaining the physical model in each residual preset time period.
Further, the predicting the operation condition of each part of the generator set through the physical model specifically includes:
acquiring a physical model closest to a current time node from the physical models in a preset time period;
and predicting the degradation position and degradation degree of each part of the generator set through a physical model closest to the current time node.
Further, the generating an adjustment scheme of each component of the generator set according to the operation condition prediction result and the performance parameter specifically includes:
judging the performance deviation condition of each part of the generator set according to the predicted degradation position and degradation degree of each part of the generator set;
and after the deviation condition is determined to be larger than the preset range, generating an adjustment scheme according to the degradation position, the degradation degree and the performance parameters of each part of the generator set.
Further, the method also comprises the following steps:
acquiring historical electricity consumption demand of a user;
predicting the electricity demand at the future moment according to the historical electricity demand of the user;
and adjusting the adjustment scheme according to the electricity demand at the future moment.
A second aspect of an embodiment of the application provides:
a performance tuning data processing system for a generator set, comprising:
the acquisition module is used for acquiring the operation condition data of each part of the generator set and calculating the performance parameters of each part of the generator set;
the model construction module is used for constructing a BP neural network model according to the collected operation condition data; constructing a physical model based on the BP neural network model;
the prediction module is used for predicting the operation condition of each part of the generator set through the physical model;
and the generating module is used for generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter.
A third aspect of embodiments of the present application provides:
a performance tuning data processing system for a generator set, comprising:
at least one memory for storing a program;
and the at least one processor is used for loading the program to execute the performance adjustment data processing method of the generator set.
A fourth aspect of an embodiment of the application provides:
a storage medium having stored therein processor-executable instructions which, when executed by a processor, are for implementing a method of processing performance tuning data for a generator set as described above.
The beneficial effects of the application are as follows: according to the application, the operation condition data of each part of the generator set is collected, the BP neural network model is constructed according to the operation condition data, then the physical model is constructed based on the constructed BP neural network model, the operation condition of each part of the generator set is predicted through the physical model, and then the adjustment scheme of each part of the generator set is generated according to the prediction result and the calculated performance parameter, so that the function of automatically adjusting the operation condition of each part of the generator set is realized according to the adjustment scheme, meanwhile, a worker can also quickly determine the position of a device to be adjusted according to the adjustment scheme, so that the adjustment and investigation time is saved, and the working efficiency of the combined cycle generator set is improved.
Drawings
FIG. 1 is a flowchart of a method for processing performance adjustment data of a generator set according to an embodiment of the present application.
Detailed Description
The application will now be described in further detail with reference to the drawings and to specific examples. The step numbers in the following embodiments are set for convenience of illustration only, and the order between the steps is not limited in any way, and the execution order of the steps in the embodiments may be adaptively adjusted according to the understanding of those skilled in the art.
Referring to fig. 1, an embodiment of the present application provides a method for processing performance adjustment data of a generator set, including the following steps:
s110, collecting operation condition data of all parts of the generator set, and calculating performance parameters of all the parts of the generator set; the method comprises the steps of collecting air state parameters of an inlet and an outlet of a gas compressor, gas state parameters of an inlet and an outlet of a turbine, gas turbine output power, gas inlet parameters of each cylinder of a steam turbine and steam turbine output power through temperature, pressure, flow and power sensors, and then respectively calculating polytropic indexes according to the temperature and the pressure parameters of the inlet and the outlet of the gas compressor and the turbine, wherein the polytropic indexes are calculated by adopting a formula 1:
wherein T is thermodynamic temperature, and the unit is K; p is absolute pressure in kPa; x is a component name abbreviation, and represents a compressor c or a gas turbine t; in is the component inlet fluid; out is the component outlet fluid; k is a polytropic index.
S120, constructing a BP neural network model according to the collected operation condition data; the method specifically comprises the step of respectively creating BP neural network models of all the years according to the operation condition data of all the years by taking the years as a unit.
S130, constructing a physical model based on the BP neural network model; the method comprises the steps of constructing a physical model by using a BP neural network model of a first year, obtaining efficiency degradation degrees of a gas compressor and a gas turbine of each operation year according to the BP neural network model of the remaining years, and correcting the physical model according to the degradation degrees to obtain the physical model conforming to different operation years. For example, in the operation year 2013-2015, taking the 2013 year as the first year, constructing a 2013-year physical model, then obtaining the efficiency degradation degree of the compressor and the gas turbine in the 2014 year by adopting the 2014-year BP neural network model, and correcting the 2013-year physical model according to the degradation degree in the 2014 year to obtain the 2014-year physical model; and then obtaining the efficiency degradation degree of the compressor and the gas turbine in 2015 by adopting the BP neural network model in 2015, and correcting the physical model in 2014 according to the degradation degree in 2015 to obtain the physical model in 2015.
S140, predicting the operation condition of each part of the generator set through the physical model; the location and extent of degradation of each component of the genset at a future time is predicted.
And S150, generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter. The performance parameter refers to the polytropic index calculated in step S110. And adjusting or replacing the working states of all parts of the generator set according to the positions and the degradation degrees of all parts of the generator set at the future time and the polytropic index, so as to ensure the stable and orderly operation of the generator set.
According to the method, the device and the system, the operation condition data of all parts of the generator set are collected, the BP neural network model is built according to the operation condition data, then the physical model is built based on the built BP neural network model, the operation condition of all parts of the generator set is predicted through the physical model, then the adjustment scheme of all parts of the generator set is generated according to the prediction result and the calculated performance parameters, so that the function of automatically adjusting the operation condition of all parts of the generator set is achieved according to the adjustment scheme, meanwhile, a worker can also quickly determine the position of a device to be adjusted according to the adjustment scheme, adjustment and investigation time is saved, and the working efficiency of the combined cycle generator set is improved.
As a preferred embodiment, the constructing a BP neural network model according to the collected operation condition data specifically includes:
storing the collected operation condition data into a subset constructed in a preset time period; the preset time period may be segmented in years. And the collected operation condition data are respectively stored in subsets of different years, so that the data can be conveniently called. Of course, the time may be segmented in units of months. Specifically, the method is adjusted according to actual needs.
Acquiring operation condition data meeting preset requirements from each subset as target data, and randomly dividing the target data into training set data and test set data; the preset requirement refers to data generated by stable operation of the generator set.
And training the neural network according to the training set data, and constructing a BP neural network model.
Before the constructed BP neural network model is applied, the reliability of the BP neural network model needs to be verified, and the verification process comprises the following steps:
acquiring test set data in the target data;
inputting the input variable of the test set data into the BP neural network model to obtain a predicted value corresponding to the input variable of the test set data;
calculating the difference value of the predicted value corresponding to the input variable and the output variable corresponding to the input variable;
and judging whether the difference value is within an engineering acceptance range, if so, judging that the reliability of the BP neural network model passes verification, otherwise, modifying training set data and re-modeling.
In some embodiments, the annual operation data of the power plant are taken as units, the collected data are effectively screened, proper sample data are selected, the power generated by the gas turbine and the power generated by the steam turbine are taken as output variables, and the coupling relation between the power generated by the gas turbine and the power generated by the steam turbine and each input variable is studied, wherein the input variables comprise T a 、P a 、m a 、ΔP in 、m f 、P 2 、P 3 、P 4 、T 2 、T 3 、T 4 、m 4 、IGV、m hp 、P hp 、P rh 、T hp 、T rh 、T fw And P c Wherein m is mass flow rate, and the unit is kg/s; IGV is the opening of the inlet adjustable guide vane; a is the environment; in is an air inlet channel of the air compressor; f is fuel; 1 is the inlet of the air compressor; 2 is the outlet of the air compressor; 3 is the inlet of the gas turbine; 4 is the outlet of the gas turbine; hp is the main steam inlet of the turbine; rh is steam turbine reheat steam; fw is the main feed water of the waste heat boiler; c is a condenser, due to turbine inlet temperature T 3 Is not easy to directly obtain, so T 3 The method is obtained through calculation of the heat balance of the waste heat boiler and the air inlet characteristic of the air compressor.
Dividing the target data into training set data and test set data;
performing neural network training on the input variable and the output variable of the training set data, and creating a BP neural network model;
the input variables of the test set are imported into a trained BP neural network to obtain the output power of the gas turbine and the predicted value of the output power of the steam turbine;
comparing the output variable of the test set with the predicted value, and judging whether the error value is in the engineering acceptance range;
and if the error value is within the engineering acceptance range, verifying the reliability of the model.
The reliability of the model is verified by the implementation, and the accuracy of the output quantity of the model is improved.
As a preferred embodiment, the building a physical model based on the BP neural network model specifically includes:
constructing a physical model according to the BP neural network model constructed in the first preset time period; the construction process is to construct an undegraded all-condition physical model by adopting a simulation platform construction system. The simulation platform is Thermoflow, ebsilon or Simulink.
Predicting the performance degradation degree of each part of the generator set in each residual preset time period through a BP neural network model constructed in the residual preset time period;
and correcting the physical model according to the performance degradation degree of each part of the generator set in each residual preset time period, and obtaining the physical model in each residual preset time period.
According to the embodiment, the physical model in each preset time period is constructed, so that the performance prediction is more intuitively and conveniently performed when the prediction is performed through the physical model.
As a preferred embodiment, the predicting, by using the physical model, the operation condition of each component of the generator set specifically includes:
acquiring a physical model closest to a current time node from the physical models in a preset time period;
and predicting the degradation position and degradation degree of each part of the generator set through a physical model closest to the current time node. The prediction result is to predict the degradation position and degradation degree of each part of the generator set at the future time.
According to the embodiment, the degradation positions and degradation degrees of all parts of the generator set are predicted through the physical model, so that corresponding measures are taken to improve the performance of the generator set.
As a preferred embodiment, the generating an adjustment scheme for each component of the generator set according to the operation condition prediction result and the performance parameter specifically includes:
judging the performance deviation condition of each part of the generator set according to the predicted degradation position and degradation degree of each part of the generator set;
and after the deviation condition is determined to be larger than the preset range, generating an adjustment scheme according to the degradation position, the degradation degree and the performance parameters of each part of the generator set. The preset range refers to the maximum error tolerance value of each component of the motor unit.
After the deviation condition is determined to be larger than the preset range, the implementation generates an adjusting scheme, so that the control end automatically adjusts the generator set or reminds a worker to perform manual intervention operation, and stable operation of the generator set is ensured.
As a preferred embodiment, the method further comprises the steps of:
acquiring historical electricity consumption demand of a user;
predicting the electricity demand at the future moment according to the historical electricity demand of the user;
and adjusting the adjustment scheme according to the electricity demand at the future moment.
In the embodiment, when the electricity consumption of the user in the future time period exceeds the maximum electricity consumption of the generator set, corresponding measures are taken to enable the electricity consumption generated by the generator set to meet the electricity consumption requirement of the user in the future time period.
In other embodiments, real-time operation parameters and calculation related performance parameters of all components of the system are acquired by using temperature, pressure, flow and power sensors arranged at all points of the gas-steam combined cycle system, the real-time operation parameters and calculation related performance parameters are uploaded to a data cloud platform through a transmission network, unit operation data under stable working conditions are screened out, a BP neural network model is built based on data analysis, and a system degradation simulation physical modeling process is carried out on the basis of the trained model. And predicting the position and the degradation degree of the future system degradation by using the established physical model, and further, determining the performance deviation caused by the system component degradation to perform unit operation diagnosis. In addition, intelligent upgrading transformation of the power plant can be performed, a suggestion and an optimization scheme are provided for the unit anti-degradation operation by combining supply and demand matching dynamic records, the generating capacity of the unit meets the requirements of external load, and the intelligent service system with man-machine interaction, comprehensive multi-level perception information, convenience and high efficiency in integration is realized.
And the BP neural network model can be constructed by the following ways:
taking annual operation data of a power plant as a unit, effectively screening the acquired data, selecting proper sample data, taking the power generated by a gas turbine and the power generated by a steam turbine as output variables, researching the coupling relation between the power generated by the gas turbine and the power generated by the steam turbine and each input variable, wherein the input variables comprise T a 、P a 、m a 、ΔP in 、m f 、P 2 、P 3 、P 4 、T 2 、T 3 、T 4 、m 4 、IGV、m hp 、P hp 、P rh 、T hp 、T rh 、T fw And P c Wherein m is mass flow rate, and the unit is kg/s; IGV is the opening of the inlet adjustable guide vane; a is the environment; in is an air inlet channel of the air compressor; f is fuel; 1 is the inlet of the air compressor; 2 is the outlet of the air compressor; 3 is the inlet of the gas turbine; 4 is the outlet of the gas turbine; hp is the main steam inlet of the turbine; rh is steam turbine reheat steam; fw is the main feed water of the waste heat boiler; c is a condenser, due to turbine inlet temperature T 3 Is not easy to directly obtain, and is obtained by calculating the heat balance of the waste heat boiler and the air inlet characteristic of the air compressor.
Then, dividing the screened 90% effective data samples into training set data, and classifying the rest 10% samples into test set data; and training the neural network by using the input variable and the output variable of the training set data, and creating a BP neural network model.
Specifically, gas turbine output P gt The corresponding BP neural network modeling process is as follows:
firstly, initializing a network;
determining the number of network input layer nodes n=12, the number of network output layer nodes m=1, and initiallyChemical connection weight omega ij 、ω jk And a hidden layer threshold a, an output layer threshold b, the optimal hidden layer node number l refers to formula 2-formula 4:
l < n-1 formula 2
l＝log 2 n formula 4
Wherein c is a constant between 0 and 10.
Secondly, outputting H by an hidden layer;
the hidden layer output H is calculated using equations 5 and 6:
wherein omega ij Representing connection weights between the input layer and hidden layer neurons; a, a j Is an implicit layer threshold; x is x i For input variables, this step is T a 、P a 、m a 、ΔP in 、m f 、P 2 、P 3 、P 4 、T 2 、T 3 、T 4 、m 4 And IGVs.
Third step, output layer O k ；
The output layer O k Calculated using equation 7:
wherein omega jk Connection weights between hidden layer and output layer neurons; b k Is the output layer threshold.
Fourth, calculating errors;
the error value is calculated using equation 8:
e k ＝Y k -O k k=1, 2 … m equation 8
Wherein e k Is an error value, O k For outputting of the output layer, Y k The actual output power of the gas turbine corresponding to each working condition point.
Fifthly, updating the weight;
the weight update uses equations 9 and 10:
ω jk ＝ω jk +ηH j e k j=1, 2 … l; k=1, 2, … m equation 10
Where η is the learning rate.
Sixth, updating the threshold value;
the threshold updating process employs equations 11 and 12:
b k ＝b k +e k k=1, 2, … m equation 12
Seventh step, judging the calculated output power P of the gas turbine gt If the error is within the error allowable range, returning to the second step for loop execution if the error is beyond the error allowable range.
Similarly, the BP neural network modeling process corresponding to the output power of the steam turbine is as follows:
firstly, initializing a network;
determining the number of network input layer nodes n=10, the number of network output layer nodes m=1, and initializing a connection weight omega ij 、ω jk And hidden layer threshold a, output layer threshold b, the optimal hidden layer node number l is referred to equation 13-equation 15:
l < n-1 equation 13
l＝log 2 n formula 15
Wherein c is a constant between 0 and 10.
Secondly, outputting H by an hidden layer;
the hidden layer output H is calculated using equations 16 and 17:
wherein omega ij Representing connection weights between the input layer and hidden layer neurons; a, a j Is an implicit layer threshold; x is x i For input variables, this step is T 4 、P 4 、m 4 、m hp 、P hp 、T hp 、P rh 、T rh 、T fw And P c 。
Third step, output layer O k ；
The output layer O k Calculated using equation 18:
wherein omega jk Connection weights between hidden layer and output layer neurons; b k Is the output layer threshold.
Fourth, calculating errors;
the error value is calculated using equation 19:
e k ＝Y k -O k k=1, 2 … m equation 19
Wherein e k Is an error value, O k For outputting of the output layer, Y k Steam turbine corresponding to each working condition pointThe actual output power.
Fifthly, updating the weight;
the weight update uses equations 20 and 21:
ω jk ＝ω jk +ηH j e k j=1, 2 … l; k=1, 2, … m equation 21
Where η is the learning rate.
Sixth, updating the threshold value;
the threshold update procedure employs equations 22 and 23:
b k ＝b k +e k k=1, 2, … m equation 23
Seventh step, judging the calculated output power P of the steam turbine gt If the error is within the error allowable range, returning to the second step for loop execution if the error is beyond the error allowable range.
After BP neural network models of all years are built, physical models of all working conditions of an undegraded system are built on a simulation platform by utilizing the BP neural network models of the first year, the efficiency degradation degree of the compressor and the gas turbine and the system output change of all the operating years are obtained on the basis of the BP neural network models of the following years, the physical models of all the years are obtained by correcting on the basis of the physical models of the first year according to the degradation degree of the components, and the all working condition operation characteristics of the undeveloped system can be predicted on the basis of the dynamic physical models.
Then, by analyzing the supply and demand matching dynamic records, whether the future power supply is matched with the expected load is judged. When the power supply can not meet the external load requirement, the system to which the generator set belongs is regulated, if the power supply and load deviation exceeds the self-regulating capacity of the system, the expert remote diagnosis and technical support are directly requested, otherwise, the system starts the equipment for early warning and self-regulating. In order to make the adjustment process more accurate, a feedback process is introduced. In the adjusting process, the parts with large degradation degree are preferentially adjusted according to the variable index change amplitude of the compressor and the turbine. The main regulation method of the gas compressor is to regulate the opening of the IGV and clean the gas compressor on line, and the main regulation method of the turbine is to regulate and change the opening of the IGV by combining fuel, and the inlet temperature is increased by readjusting the exhaust temperature of the turbine so as to maintain higher output. When turbine exhaust parameters are varied, the steam parameters are adjusted to obtain greater output using the relationship of the steam turbine output power to its associated input variables. And the operation is repeated until the output power of the system meets the external load requirement.
The embodiment of the application also provides a system for processing the performance adjustment data of the generator set corresponding to the method of fig. 1, which comprises the following steps:
the acquisition module is used for acquiring the operation condition data of each part of the generator set and calculating the performance parameters of each part of the generator set;
the model construction module is used for constructing a BP neural network model according to the collected operation condition data; constructing a physical model based on the BP neural network model;
the prediction module is used for predicting the operation condition of each part of the generator set through the physical model;
and the generating module is used for generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter.
The content of the method embodiment of the application is suitable for the system embodiment, the specific function of the system embodiment is the same as that of the method embodiment, and the achieved beneficial effects are the same as those of the method.
The embodiment of the application also provides a system for processing the performance adjustment data of the generator set, which comprises the following steps:
at least one memory for storing a program;
and the at least one processor is used for loading the program to execute the performance adjustment data processing method of the generator set.
The content of the method embodiment of the application is suitable for the system embodiment, the specific function of the system embodiment is the same as that of the method embodiment, and the achieved beneficial effects are the same as those of the method.
In addition, the embodiment of the application further provides a storage medium, wherein the storage medium stores processor-executable instructions, and the processor-executable instructions are used for realizing the performance adjustment data processing method of the generator set when being executed by a processor.
In summary, in the embodiment of the method, the operation condition data of each component of the generator set is collected, the BP neural network model is built according to the operation condition data, then the physical model is built based on the built BP neural network model, the operation condition of each component of the generator set is predicted through the physical model, and then the adjustment scheme of each component of the generator set is generated according to the prediction result and the calculated performance parameter, so that the function of automatically adjusting the operation condition of each component of the generator set is realized according to the adjustment scheme, meanwhile, a worker can also quickly determine the position of a device to be adjusted according to the adjustment scheme, so that adjustment and investigation time is saved, and the working efficiency of the combined cycle generator set is improved.
While the preferred embodiment of the present application has been described in detail, the present application is not limited to the embodiments described above, and various equivalent modifications and substitutions can be made by those skilled in the art without departing from the spirit of the present application, and these equivalent modifications and substitutions are intended to be included in the scope of the present application as defined in the appended claims.
Claims (8)
1. A method for processing performance adjustment data of a generator set is characterized by comprising the following steps of: the method comprises the following steps:
collecting operation condition data of each part of the generator set, and calculating performance parameters of each part of the generator set, wherein the performance parameters are polytropic indexes;
constructing a BP neural network model according to the collected operation condition data;
constructing a physical model according to the BP neural network model constructed in the first preset time period;
predicting the performance degradation degree of each part of the generator set in each residual preset time period through a BP neural network model constructed in the residual preset time period;
correcting the physical model according to the performance degradation degree of each part of the generator set in each residual preset time period, and obtaining the physical model in each residual preset time period in sequence;
acquiring a physical model closest to a current time node from the physical models in a preset time period;
performing operation condition prediction on each part of the generator set through a physical model closest to the current time node, wherein the operation condition prediction comprises degradation position prediction and degradation degree prediction;
and generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter.
2. The method for processing performance adjustment data of a generator set according to claim 1, wherein: the BP neural network model is constructed according to the collected operation condition data, and the BP neural network model specifically comprises:
storing the collected operation condition data into a subset constructed in a preset time period;
acquiring operation condition data meeting preset requirements from each subset as target data;
randomly dividing the target data into training set data and test set data;
and training the neural network according to the training set data, and constructing a BP neural network model.
3. The method for processing performance adjustment data of a generator set according to claim 2, wherein: the BP neural network model is also corrected by:
acquiring test set data in the target data;
inputting the input variable of the test set data into the BP neural network model to obtain a predicted value corresponding to the input variable of the test set data;
calculating the difference value of the predicted value corresponding to the input variable and the output variable corresponding to the input variable;
and judging whether the difference value is within an engineering acceptance range, if so, judging that the reliability of the BP neural network model passes verification, otherwise, modifying training set data and re-modeling.
4. The method for processing performance adjustment data of a generator set according to claim 1, wherein: and generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter, wherein the adjusting scheme specifically comprises the following steps of:
judging the performance deviation condition of each part of the generator set according to the predicted degradation position and degradation degree of each part of the generator set;
and after the deviation condition is determined to be larger than the preset range, generating an adjustment scheme according to the degradation position, the degradation degree and the performance parameters of each part of the generator set.
5. The method for processing performance adjustment data of a generator set according to claim 4, wherein: the method also comprises the following steps:
acquiring historical electricity consumption demand of a user;
predicting the electricity demand at the future moment according to the historical electricity demand of the user;
and adjusting the adjustment scheme according to the electricity demand at the future moment.
6. A performance adjustment data processing system for a generator set, comprising: comprising the following steps:
the acquisition module is used for acquiring the operation condition data of each part of the generator set and calculating the performance parameters of each part of the generator set;
the model construction module is used for constructing a BP neural network model according to the collected operation condition data; constructing a physical model according to the BP neural network model constructed in the first preset time period; predicting the performance degradation degree of each part of the generator set in each residual preset time period through a BP neural network model constructed in the residual preset time period; correcting the physical model according to the performance degradation degree of each part of the generator set in each residual preset time period, and obtaining the physical model in each residual preset time period in sequence;
the prediction module is used for acquiring a physical model closest to the current time node from the physical models in a preset time period; performing operation condition prediction on each part of the generator set through a physical model closest to the current time node, wherein the operation condition prediction comprises degradation position prediction and degradation degree prediction;
and the generating module is used for generating an adjusting scheme of each part of the generator set according to the operation condition prediction result and the performance parameter.
7. A performance adjustment data processing system for a generator set, comprising: comprising the following steps:
at least one memory for storing a program;
at least one processor for loading the program to perform a method of performance tuning data processing of a generator set according to any one of claims 1-5.
8. A storage medium having stored therein instructions executable by a processor, characterized by: the processor-executable instructions, when executed by a processor, are for implementing a method of performance adjustment data processing of a generator set according to any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010057414.3A CN111275320B (en) | 2020-01-19 | 2020-01-19 | Performance adjustment data processing method, system and storage medium of generator set |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010057414.3A CN111275320B (en) | 2020-01-19 | 2020-01-19 | Performance adjustment data processing method, system and storage medium of generator set |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111275320A CN111275320A (en) | 2020-06-12 |
| CN111275320B true CN111275320B (en) | 2023-08-25 |
Family
ID=70998745
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010057414.3A Active CN111275320B (en) | 2020-01-19 | 2020-01-19 | Performance adjustment data processing method, system and storage medium of generator set |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111275320B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114135397B (en) * | 2021-11-05 | 2023-04-25 | 广州珠江天然气发电有限公司 | Diagnosis optimization method, device and system for gas-steam combined cycle performance |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101487466A (en) * | 2009-02-25 | 2009-07-22 | 华东理工大学 | On-line soft measuring method for compression ratio and polytropic efficiency of centrifugal compressor |
| CN104656443A (en) * | 2014-12-31 | 2015-05-27 | 重庆邮电大学 | HCCI engine ignition timing self-adaptive PID control method based on BP neural network |
| WO2019036095A1 (en) * | 2017-08-17 | 2019-02-21 | Siemens Aktiengesellschaft | Deep convolutional neural network based anomaly detection for transactive energy systems |
| CN109711614A (en) * | 2018-12-24 | 2019-05-03 | 新奥数能科技有限公司 | A kind of the dynamic optimization progress control method and system of distributed busbar protection |
| CN110056544A (en) * | 2019-03-12 | 2019-07-26 | 马文德 | A kind of compressor most preferably washes period acquisition methods |
-
2020
- 2020-01-19 CN CN202010057414.3A patent/CN111275320B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101487466A (en) * | 2009-02-25 | 2009-07-22 | 华东理工大学 | On-line soft measuring method for compression ratio and polytropic efficiency of centrifugal compressor |
| CN104656443A (en) * | 2014-12-31 | 2015-05-27 | 重庆邮电大学 | HCCI engine ignition timing self-adaptive PID control method based on BP neural network |
| WO2019036095A1 (en) * | 2017-08-17 | 2019-02-21 | Siemens Aktiengesellschaft | Deep convolutional neural network based anomaly detection for transactive energy systems |
| CN109711614A (en) * | 2018-12-24 | 2019-05-03 | 新奥数能科技有限公司 | A kind of the dynamic optimization progress control method and system of distributed busbar protection |
| CN110056544A (en) * | 2019-03-12 | 2019-07-26 | 马文德 | A kind of compressor most preferably washes period acquisition methods |
Non-Patent Citations (1)
| Title |
|---|
| 杨承 等.联合循环机组中燃气轮机部件劣化 对功率效率的敏感性分析.中 国 电 机 工 程 学 报.2015,第35卷(第35期),第5803-5810页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111275320A (en) | 2020-06-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Gas turbine performance prediction via machine learning | |
| US11436395B2 (en) | Method for prediction of key performance parameter of an aero-engine transition state acceleration process based on space reconstruction | |
| Tsoutsanis et al. | A component map tuning method for performance prediction and diagnostics of gas turbine compressors | |
| US10732595B2 (en) | Control parameter automatic-adjustment apparatus, control parameter automatic-adjustment method, and control parameter automatic-adjustment apparatus network | |
| CN110863919B (en) | Engine virtual test environment system and engine management system mapping method | |
| CN103732895A (en) | Sliding mode controller, and internal combustion engine system control apparatus | |
| US9952566B2 (en) | Method for controlling and/or regulating a technical system in a computer-assisted manner | |
| Tsoutsanis et al. | An efficient component map generation method for prediction of gas turbine performance | |
| CN110837223A (en) | Combustion optimization control method and system for gas turbine | |
| US11939931B2 (en) | Engine control system | |
| CN111275320B (en) | Performance adjustment data processing method, system and storage medium of generator set | |
| CN113795660A (en) | Internal combustion engine controller | |
| US9423781B2 (en) | Model based control with engine perturbation feedback | |
| de Nola et al. | A Model-Based Computer Aided Calibration Methodology Enhancing Accuracy, Time and Experimental Effort Savings Through Regression Techniques and Neural Networks | |
| CN110991073B (en) | Simulation method, device and equipment for liquid rocket engine | |
| CN112782974B (en) | Control parameter calibration method | |
| CN114861126A (en) | Method and system for predicting stability in combustion adjustment process of gas turbine | |
| CN111894752B (en) | Model predictive control algorithm-based diesel engine VGT-EGR control method | |
| CN114810337A (en) | Control method and related device of supercharger | |
| US20220235721A1 (en) | Internal combustion engine controller | |
| Salehi et al. | Selection and tuning of a reduced parameter set for a turbocharged diesel engine model | |
| CN116517867B (en) | Method and device for diagnosing and suppressing surge of compressor | |
| Huan et al. | Airpath prediction model and predictive control of turbocharged gasoline engine. | |
| Yang et al. | On-Line Component Map Adaptive Procedure Based on Sensor Data | |
| Nikzadfar et al. | Optimal idle speed control of a light duty turbodiesel engine with the aim of minimizing fuel consumption |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |