CN112909936B - Method, device and system for monitoring running state of thermal power generating unit - Google Patents
Method, device and system for monitoring running state of thermal power generating unit Download PDFInfo
- Publication number
- CN112909936B CN112909936B CN202110155558.7A CN202110155558A CN112909936B CN 112909936 B CN112909936 B CN 112909936B CN 202110155558 A CN202110155558 A CN 202110155558A CN 112909936 B CN112909936 B CN 112909936B
- Authority
- CN
- China
- Prior art keywords
- related parameters
- thermal power
- historical
- generating unit
- load related
- 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
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
-
- 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
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- 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 invention discloses a method, a device and a system for monitoring the running state of a thermal power generating unit, which are used for solving the problem of inaccurate running state monitoring of the thermal power generating unit. This scheme includes: acquiring historical load related parameters and actual load related parameters of the thermal power generating unit before a first historical time period; inputting part of historical load related parameters into a matched parameter model of the thermal power generating unit to obtain predicted load related parameters; taking the historical load related parameters as sample input parameters, and taking the predicted load related parameters and the actual load related parameters as sample output parameters to train a parameter model; and monitoring load related parameters of the thermal power generating unit, inputting the load related parameters into the trained parameter model, and determining the operating state of the thermal power generating unit according to an output result of the parameter model. In the scheme, the samples for training the parameter model are generated by the actual parameters and the prediction parameters together, so that the training process has inheritance, the reliability of the parameter model is improved, and the monitoring accuracy of the running state of the thermal power generating unit is improved.
Description
Technical Field
The invention relates to the field of state monitoring, in particular to a method, a device and a system for monitoring the running state of a thermal power generating unit.
Background
With the development of the intelligent technology of the coal-fired thermal power generating unit, a large number of deep learning algorithms are widely applied to the thermal power generating unit. The neural network model can be used for predicting the output parameters of the thermal power generating unit, further monitoring the running state of the unit according to the predicted parameters, and generating early warning information when the state is abnormal. However, with the long-term operation of the unit equipment, due to system modification, equipment maintenance or equipment performance degradation, the mapping relationship between the input parameters and the output parameters of the thermal power generating unit changes, and further the neural network model mismatch is caused. The output parameters of the thermal power generating unit cannot be accurately predicted by the mismatched neural network model, and the running state of the unit cannot be correctly judged.
For mismatched models, it is often computationally expensive and time consuming to retrain using historical parameters. How to improve the monitoring accuracy of thermal power unit running state is the technical problem that this application will solve.
Disclosure of Invention
The embodiment of the application aims to provide a method, a device and a system for monitoring the operating state of a thermal power generating unit, and aims to solve the problem that the monitoring of the operating state of the thermal power generating unit is inaccurate.
In a first aspect, a method for monitoring an operating state of a thermal power generating unit is provided, which includes:
acquiring historical load related parameters of a thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period, wherein the load related parameters comprise power consumption parameters of the thermal power generating unit and/or equipment operation physical parameters of the thermal power generating unit;
inputting part of the historical load related parameters into a parameter model matched with the thermal power generating unit to obtain predicted load related parameters of the thermal power generating unit predicted by the parameter model in the first historical time period;
taking the historical load related parameters as sample input parameters, and taking the predicted load related parameters and the actual load related parameters as sample output parameters to train the parameter model;
and monitoring load related parameters of the thermal power generating unit, inputting the load related parameters into the trained parameter model, and determining the operating state of the thermal power generating unit according to the output result of the parameter model.
In a second aspect, a device for monitoring an operating state of a thermal power generating unit is provided, which includes:
the system comprises an acquisition module, a storage module and a processing module, wherein the acquisition module is used for acquiring historical load related parameters of a thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period, and the load related parameters comprise power consumption parameters of the thermal power generating unit and/or equipment operation physical parameters of the thermal power generating unit;
the prediction module is used for inputting part of the historical load related parameters into a matched parameter model of the thermal power generating unit so as to obtain the predicted load related parameters of the thermal power generating unit in the first historical time period, which are predicted by the parameter model;
the training module is used for training the parameter model by taking the historical load related parameters as sample input parameters and taking the predicted load related parameters and the actual load related parameters as sample output parameters;
and the monitoring module monitors load related parameters of the thermal power generating unit, inputs the load related parameters into the trained parameter model, and determines the running state of the thermal power generating unit according to the output result of the parameter model.
In a third aspect, a system for monitoring an operating state of a thermal power generating unit includes:
a thermal power generating unit;
the monitoring device for the operating state of the thermal power generating unit according to the second aspect is in communication connection with the thermal power generating unit.
In a fourth aspect, an electronic device is provided, the electronic device comprising a processor, a memory and a computer program stored on the memory and executable on the processor, the computer program, when executed by the processor, implementing the steps of the method according to the first aspect.
In a fifth aspect, a computer-readable storage medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, realizes the steps of the method as in the first aspect.
In the embodiment of the application, historical load related parameters, predicted output parameters and historical output parameters of the thermal power generating unit are obtained; training a parameter model according to the historical load related parameters, the prediction output parameters and the historical output parameters; monitoring real-time operation parameters and real-time output parameters of the thermal power generating unit; and inputting the real-time operation parameters into the trained parameter model, and determining the operation state of the thermal power generating unit according to the output parameters and the real-time output parameters of the trained parameter model. The parameters used for training the parameter model are generated based on the historical output parameters and the prediction output parameters, the parameter model training process can have inheritance, the reliability of the parameter model is improved, the trained parameter model can predict the output parameters of the thermal power generating unit according to the real-time operation parameters, and the accuracy of the monitored operation state is improved by combining the real-time output parameters.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention and do not limit the invention. In the drawings:
fig. 1 is one of schematic flow charts of a method for monitoring an operating state of a thermal power generating unit according to the present invention.
Fig. 2 is a second schematic flow chart of a method for monitoring an operating state of a thermal power generating unit according to the present invention.
Fig. 3 is a third schematic flow chart of a method for monitoring an operating state of a thermal power generating unit according to the present invention.
Fig. 4 is a fourth schematic flowchart of a method for monitoring an operating state of a thermal power generating unit according to the present invention.
Fig. 5 is a fifth schematic flow chart of a method for monitoring an operating state of a thermal power generating unit according to the present invention.
Fig. 6a is a sixth schematic flow chart of a method for monitoring an operating state of a thermal power generating unit according to the present invention.
Fig. 6b is a schematic diagram illustrating a generation flow of training sample data of the method for monitoring the operating state of the thermal power generating unit according to the present invention.
Fig. 7a is one of schematic structural diagrams of a monitoring device for an operating state of a thermal power generating unit according to the present invention.
Fig. 7b is a second schematic structural diagram of a monitoring device for an operating state of a thermal power generating unit according to the present invention.
Fig. 8 is a schematic structural diagram of a monitoring system for an operating state of a thermal power generating unit according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention. The reference numbers in the present application are only used for distinguishing the steps in the scheme and are not used for limiting the execution sequence of the steps, and the specific execution sequence is described in the specification.
In order to solve the problems in the prior art, an embodiment of the present application provides a method for monitoring an operating state of a thermal power generating unit, as shown in fig. 1, including the following steps:
s11: the method comprises the steps of obtaining historical load related parameters of a thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period, wherein the load related parameters comprise power consumption parameters of the thermal power generating unit and/or equipment operation physical parameters of the thermal power generating unit.
The load related parameters of the thermal power generating unit may include one or more parameters, and the power consumption parameters of the thermal power generating unit include, for example, current parameters, voltage parameters, power parameters, and other parameters related to the power of the thermal power generating unit. The power consumption parameters can represent the power consumption of the thermal power generating unit in operation, and further represent the load of the thermal power generating unit.
The device operation physical parameters of the thermal power generating unit include, for example, temperature parameters and pressure parameters of a medium in the thermal power generating unit, and may also include vibration parameters of main devices or auxiliary devices of the thermal power generating unit, and the device operation physical parameters include physical parameters that change along with load changes of the thermal power generating unit. For example, when the load of the unit is increased, the temperature of the medium in the unit is increased, and the pressure is increased. The physical parameters of the operation of the equipment can also represent the load of the live generator set.
The historical load related parameter acquired in this step may be a load related parameter of the thermal power generating unit in a historical period before the first historical period. The time period corresponding to the historical load related parameter is before the first historical time period, and the time period corresponding to the historical load related parameter may be longer, shorter or equal in length than the first historical time period. For example, assume that the first history period is a day 6: 00-7: 00, the historical load related parameter obtained in the step may be a load related parameter of the thermal power generating unit in an hour of 5.
Optionally, the first history time period is connected to the time period corresponding to the historical load related parameter, that is, the start time of the first history time period coincides with the end time of the time period corresponding to the historical load related parameter.
The historical load related parameters obtained in this step can represent the operating load of the thermal power generating unit before the first historical time period, for example, the thermal power generating unit operates at a stable load, or the load of the thermal power generating unit gradually increases with time. Under a normal condition, the operation load of the thermal power generating unit which normally operates cannot suddenly change in a short time, and the operation load usually has certain continuity. The historical load related parameters before the first historical period can characterize the load trend of the thermal power generating unit in the first historical period to a certain extent.
S12: and inputting part of the historical load related parameters into a parameter model matched with the thermal power generating unit to obtain the predicted load related parameters of the thermal power generating unit in the first historical time period, which are predicted by the parameter model.
For convenience of explanation, in this example, the historical load related parameters are divided into two groups, i.e., a group a and a group B, where the group a is a part of the historical load related parameters of the parameter model input to the thermal power generating unit. Optionally, the number of parameters included in the group a of historical load parameters is equal to the number of parameters included in the group B of load-related parameters.
The parameter model matched with the thermal power generating unit can be a parameter model obtained by training by taking the load parameter of the thermal power generating unit as a sample based on a deep learning algorithm. The parameter model is used for predicting load related parameters of the thermal power generating unit in a certain period of time in the future according to the input load related parameters.
Because the historical load related parameters are load related parameters of the thermal power generating unit in a certain period before the first historical period, the parameter model can perform prediction according to the input A group of parameters and output the predicted load related parameters of the thermal power generating unit in the first historical period.
S13: and taking the historical load related parameters as sample input parameters, and taking the predicted load related parameters and the actual load related parameters as sample output parameters to train the parameter model.
In this step, the sample input parameters and the sample output parameters may be associated into one-to-one corresponding parameter sets to train the parametric model. Because the sample input parameters and the sample output parameters for training comprise the real historical parameters of the thermal power generating unit and the prediction parameters obtained by predicting the parameter model, the training of the parameter model by using the sample input parameters and the sample output parameters can ensure that the training process has certain inheritance.
On one hand, the real historical parameters are used for training, so that the matching degree of the parameter model obtained by training and the thermal power generating unit can be improved, and the prediction accuracy is improved. On the other hand, the parameter training parameter model predicted by the parameter model has certain inheritance, so that the reliability of model training can be effectively improved, and the prediction accuracy of the trained model is further improved.
S14: and monitoring load related parameters of the thermal power generating unit, inputting the load related parameters into the trained parameter model, and determining the operating state of the thermal power generating unit according to the output result of the parameter model.
The trained parameter model can predict the load related parameters of the thermal power generating unit in the future period according to the input load related parameters. In the step, the monitored load related parameters of the thermal power generating unit are input into a parameter model obtained through training, and the output result of the parameter model can represent the load related parameters of the thermal power generating unit which may appear in a future time period, so that the running state of the thermal power generating unit is determined.
The output result of the parameter model can be used for determining the operating state of the thermal power generating unit in the monitoring period, and can also be used for assisting in judging the operating state of the thermal power generating unit in the future period.
For example, if the output result of the parameter model represents that the load of the thermal power generating unit is too high in a future period, it can be determined that the thermal power generating unit is in an operation state with increased load or large load fluctuation in a monitoring period, and then it can be determined that the operation state of the thermal power generating unit in the monitoring period is abnormal, adjustment needs to be performed on the thermal power generating unit, and the load superelevation in the future period is avoided.
For another example, the load threshold of the thermal power generating unit in the future period may be generated according to the output result of the parametric model. Assuming that the output result of the parameter model represents that the load of the thermal power generating unit in the future time period is 40% -60%, the load threshold value of the thermal power generating unit in the corresponding future time period can be set to be 40% -60%, and the actual load related parameters of the thermal power generating unit are continuously monitored. If the actual load related parameters of the thermal power generating unit exceed the set load threshold value in the future time period, it is determined that the thermal power generating unit does not operate according to the expected operation state, namely the operation state of the thermal power generating unit is abnormal, and the reason for the abnormality needs to be checked in time so as to ensure the operation safety and economy of the thermal power generating unit.
Based on the solution provided by the foregoing embodiment, optionally, before the foregoing step S11, as shown in fig. 2, the method further includes:
s21: and acquiring actual load related parameters of the thermal power generating unit in a second historical time period and predicted load related parameters of the thermal power generating unit in the second historical time period, which are predicted by the parameter model.
The second history period in this step may have at least a partially overlapped period with the first history period, or may be two periods that are completely non-overlapped. Before this step, the parameter model may predict, according to the load-related parameter before the second history period, the predicted load-related parameter of the thermal power generating unit in the second history period. The actual load related parameters of the thermal power generating unit in the second historical time period can be obtained through monitoring of the sensor.
S22: and judging whether the parameter models are mismatched according to the similarity of the actual load related parameters in the second historical period and the predicted load related parameters in the second historical period.
Specifically, the similarity in this step may be calculated by using a calculation method of euclidean distance and/or cosine similarity, and whether the parameter model is mismatched or not may be determined according to the magnitude or variation trend of the similarity. For example, in the second history period, the similarity between the actual load-related parameter and the predicted load-related parameter gradually decreases with time, and the parameter model mismatch is determined. For another example, if at least one similarity value in the second history period is lower than the preset similarity value, the parameter model mismatch is determined.
According to the scheme provided by the embodiment of the application, whether the parameter model can correctly predict the load related parameters of the thermal power unit in the future time period or not can be judged according to whether the real load related parameters of the thermal power unit in the second historical time period are similar to the load related parameters of the thermal power unit in the second historical time period predicted by the parameter model, and whether the parameter model is mismatched or not can be accurately judged.
Wherein, the step S11 includes:
s23: when the parameter models are mismatched, acquiring historical load related parameters of the thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period.
Through the steps, whether the parameter model is mismatched or not can be judged, and the acquisition of the load related parameters is executed during the mismatch, so that the parameter model is trained to improve the matching degree of the parameter model and the thermal power generating unit, and the trained parameter model can accurately predict the load related parameters of the thermal power generating unit in the future period.
If the parameter model is not mismatched, the parameter model can accurately predict the load related parameters of the thermal power generating unit in the future period. The load related parameters of the thermal power generating unit can be directly monitored by using the unmatched parameter model, and the operating state of the thermal power generating unit can be determined according to the output result of the parameter model.
Based on the solution provided by the foregoing embodiment, optionally, as shown in fig. 3, the foregoing step S21 includes:
s31: acquiring an actual load parameter set X of the thermal power generating unit in a second historical period i ＝[x 1 ,x 2 ，…,x n ]And a set of model prediction parameters
Wherein n is a positive integer greater than 1, and the actual load parameter set X i And the model prediction parameter set
Each item in (1) corresponds to each other one by one;
the set of actual load parameters may comprise actual load parameters collected by the sensor at different times during the second history period. For example, the sensors may uniformly collect actual load parameters over the second historical period at 1 minute intervals. Assuming that the time length of the second history period is 10 minutes, 10 actual load parameters may be collected at 1 minute intervals, and the actual load parameter set may be generated based on the 10 actual load parameters. Further, the actual load parameters may be arranged in time sequence according to the time when the actual load parameters are collected, so as to generate the actual load parameter group arranged in time sequence.
The set of model prediction parameters may include prediction parameters corresponding to a second historical period of the parametric model prediction output. And each item in the actual load parameter set corresponds to each item in the model prediction parameter set on a one-to-one basis on time. If the actual load parameters are acquired at intervals of 1 minute, the parameter model predicts the load-related parameters at the moment of acquiring the actual load parameters, and when the time length of the second history period is 10 minutes, an actual load parameter set containing 10 actual load parameters and a model prediction parameter set containing 10 model prediction parameters can be obtained through the step. The time corresponding to each item in the actual load parameter set is the same as the time corresponding to each item in the model prediction parameter set.
Wherein, the step S22 includes:
s32: respectively determining the actual load parameter set X i And the corresponding set of model prediction parameters
Similarity of each item in (1)
Wherein dot represents the dot product of two vectors; | | represents the two-norm of the vector; sigma is a set value which can be manually set according to actual needs. Taking the above-mentioned actual load parameter set including 10 actual load parameters and the model prediction parameter set including 10 model prediction parameters as an example, the similarity of each group of parameters is calculated based on the corresponding time, so as to obtain the similarity of 10 groups based on time. These 10 similarities represent the similarity between the actual load parameter and the model prediction parameter at the corresponding time.
S33: comparing the similarity based on time sequence
Dividing the obtained object into k groups, and determining similarity mean A of each group vgk Wherein k is a positive integer less than or equal to n.
In this step, the similarity may be sorted based on the time corresponding to the multiple similarities to obtain a time-based similarity array. The sorted similarities are then divided into k groups. Optionally, each group contains the same number of similarities. For example, assuming that k is 5, 10 chronological similarities are divided into 5 groups in this step, each group including 2 similarities.
Then, the mean value A of each group of similarity is respectively obtained vgk Namely:
s34: and determining the parameter model mismatch when the similarity mean value meets a mismatch standard, wherein the mismatch standard comprises that k groups of similarity mean values arranged based on time sequence are sequentially reduced, and/or at least one group of similarity mean values in the k groups of similarity mean values are smaller than a preset similarity mean value.
In this example, if A vg1 >A vg2 >A vg3 >A vg4 >A vg5 If yes, the similarity is considered to be in a gradual descending trend, and model mismatch can be determined. Alternatively, there is at least one mean value of similarity A vgk And if the similarity is smaller than a certain fixed preset similarity threshold, determining the model mismatch.
By the scheme provided by the embodiment of the application, whether the parameter model is mismatched or not is judged based on the similarity of the actual load parameter and the model prediction parameter, whether the parameter model can correctly predict the load related parameter or not can be accurately judged, and the state of the parameter model can be accurately identified.
Based on the solution provided by the foregoing embodiment, optionally, as shown in fig. 4, the foregoing step S11 includes:
s41: n historical load related parameters of the thermal power generating unit before a first historical time period are obtained.
After determining the mismatch of the parametric model, the parametric model may be updated and trained through the scheme provided by this example. In order to improve the training effect and optimize the quality of the parameter model obtained by training, the scheme provided by the embodiment processes and screens the sample data used for updating the training.
The N historical load-related parameters acquired in this step may be load-related parameters having continuity over time. Such as load related parameters collected at regular time intervals within a history period prior to the first history period.
S42: and determining extreme values of the N historical load-related parameters.
The extreme values of the historical load-related parameter may include a maximum value (an upper limit value max) and a minimum value (a lower limit value min), and the extreme values may represent the range of values max to min of the historical load-related parameter.
S43: and equally dividing the N historical load related parameters into m groups according to the extreme values of the historical load related parameters, wherein m is an integer greater than 1.
In this step, the historical load value-related parameter intervals [ min, max ] may be generated according to the extreme values, and then the intervals may be divided into m groups. Optionally, the interval is equally divided, so that the parameter intervals included in each group have the same width, that is, the interval width of each share is (max-min)/m.
S44: determining distribution density of each group of historical load related parameters
Wherein k is i Is the total number of samples in the ith group, i is an integer, and i is more than or equal to 1 and less than or equal to m.
In this step, the distribution densities of the historical load related parameters of each group are respectively calculated, and each determined distribution density can represent the distribution condition of the historical load related parameters on the interval.
S45: and screening the historical load related parameters according to the distribution density of the historical load related parameters to obtain the screened historical load related parameters meeting a preset density standard, wherein the preset density standard comprises that the difference of the distribution density of the historical load related parameters of each group is smaller than a preset distribution density difference.
In the step, the historical load related parameters are screened to obtain the uniformly distributed historical load related parameters.
Wherein, the step S12 includes:
s46: and inputting part of the screened historical load related parameters into a matched parameter model of the thermal power generating unit to obtain the predicted load related parameters of the thermal power generating unit in the first historical time period, which are predicted by the parameter model.
The filtered historical load related parameters are uniformly distributed in the interval, so that the prediction accuracy of the parameter model in each region value can be comprehensively evaluated. And further ensuring the comprehensiveness of the subsequently generated samples, thereby ensuring that the parameter model obtained by training can accurately predict the load related parameters under various working conditions, and effectively improving the quality of the parameter model obtained by training.
Based on the solution provided by the foregoing embodiment, optionally, as shown in fig. 5, step S46 includes:
s51: and obtaining sample load parameters from the filtered historical load related parameters, wherein the sample load parameters comprise part of the historical load related parameters in each group of the historical load related parameters.
In this step, the historical load related parameters obtained by screening include at least one historical load related parameter in each group, and the historical load related parameters obtained by screening are uniformly distributed in the parameter interval, so that the distribution uniformity of the historical load related parameters obtained by screening in the parameter interval is further improved.
S52: and inputting the sample load parameters into the parameter model to obtain the predicted load related parameters of the thermal power generating unit predicted by the parameter model in the first historical time period.
According to the scheme provided by the embodiment of the application, the acquired sample load parameters comprise various groups of historical load related parameters, and the sample load parameters are uniformly distributed in the parameter interval, so that the prediction accuracy of the trained parameter model on the load related parameters under various working conditions can be improved.
Based on the solution provided by the foregoing embodiment, optionally, as shown in fig. 6a, step S13 includes:
s61: and generating sample input parameters according to the historical load related parameters.
Assuming that there are 10 historical load-related parameters, some or all of the historical load-related parameters may be used as samples to generate sample input parameters. In this example, it is assumed that all of the 10 historical load-related parameters are used as samples, and 10 sample input parameters are generated.
S62: and generating sample output parameters corresponding to the historical load related parameters one by one according to the predicted load related parameters and the actual load related parameters.
In this example, it is assumed that 5 of the 10 historical load-related parameters are input to the parameter model as the group a load-related parameters, and the output results in 5 predicted load-related parameters. Then, these 5 predicted load-related parameters are used as sample output parameters, and correspond to the a-group load-related parameters one-to-one. In addition, the historical load related parameters also include 5 load related parameters of the parameter model which is not input, the 5 load related parameters of the parameter model which is not input are used as a group B, and 5 corresponding parameters are selected from the actual load related parameters and are in one-to-one correspondence with the parameters in the group B. That is, the sample output parameters generated in this step include 5 predicted load-related parameters corresponding to the group a parameters and 5 actual load-related parameters corresponding to the group B parameters.
S63: and training the parameter model according to the sample input parameters and the sample output parameters.
Because the sample output parameters generated in the steps comprise the actual parameter values and the parameter values predicted by the model, the parameter model is trained by utilizing the sample output parameters and the corresponding sample input parameters, so that the parameter model has certain inheritance in the training process, the quality of the trained parameter model is improved, and the reliability of the parameter model is improved.
Optionally, fig. 6b shows a generation process of training sample data including a sample input parameter and a sample output parameter. The method comprises the steps of firstly inputting historical input data into a neural network model, carrying out similarity comparison calculation on a result output by the model and an actual output result of the thermal power generating unit, analyzing the degradation trend of a parameter model according to the similarity, and further judging whether the parameter model is mismatched. And if the parameter models are mismatched, acquiring original data of the thermal power generating unit in a certain historical time period, and screening the original data through data distribution density calculation to obtain uniformly distributed original data. And then dividing the screened data into two parts equally, inputting one part of the data into a parameter model to obtain a predicted load related parameter, and combining the predicted load related parameter with the other part of the screened data to obtain a sample output parameter for obtaining training sample data.
The parameters used for training the parameter model are generated based on the historical output parameters and the prediction output parameters, the parameter model training process can have inheritance, the reliability of the parameter model is improved, the trained parameter model can predict the output parameters of the thermal power generating unit according to the real-time operation parameters, and the accuracy of the monitored operation state is improved by combining the real-time output parameters.
In order to solve the problems existing in the prior art, an embodiment of the present application further provides a monitoring device 70 for an operating state of a thermal power generating unit, as shown in fig. 7a, including:
the obtaining module 71 is configured to obtain historical load related parameters of a thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period, where the load related parameters include a power consumption parameter of the thermal power generating unit and/or an equipment operation physical parameter of the thermal power generating unit;
the prediction module 72 is configured to input a part of the historical load related parameters into a parameter model matched with the thermal power generating unit, so as to obtain predicted load related parameters of the thermal power generating unit predicted by the parameter model in the first historical time period;
a training module 73, which takes the historical load-related parameters as sample input parameters, and takes the predicted load-related parameters and the actual load-related parameters as sample output parameters to train the parameter model;
and the monitoring module 74 is used for monitoring the load related parameters of the thermal power generating unit, inputting the load related parameters into the trained parameter model, and determining the operating state of the thermal power generating unit according to the output result of the parameter model.
Based on the apparatus provided in the foregoing embodiment, optionally, as shown in fig. 7b, the apparatus further includes a determining module 75, configured to, before acquiring the historical load related parameter of the thermal power generating unit before the first historical period and the actual load related parameter of the thermal power generating unit in the first historical period:
acquiring actual load related parameters of the thermal power generating unit in a second historical time period and predicted load related parameters of the thermal power generating unit in the second historical time period, which are predicted by the parameter model;
judging whether the parameter model is mismatched according to the similarity of the actual load related parameters in the second historical period and the predicted load related parameters in the second historical period;
wherein the obtaining module 71 is configured to:
when the parameter models are mismatched, acquiring historical load related parameters of the thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period.
Based on the apparatus provided in the foregoing embodiment, optionally, the determining module 75 is configured to:
acquiring an actual load parameter set X of the thermal power generating unit in a second historical period i ＝[x 1 ,x 2 ，…,x n ]And a set of model prediction parameters
Wherein n is a positive integer greater than 1, and the actual load parameter set X i And the model prediction parameter set
Each item in (1) corresponds to each other one by one;
respectively determining the actual load parameter sets X i And the corresponding model prediction parameter set
Similarity of each item in (1)
Determining the similarity based on a time sequence
Dividing the obtained object into k groups, and determining similarity mean A of each group vgk Wherein k is a positive integer less than or equal to n;
and determining the parameter model mismatch when the similarity mean value meets a mismatch standard, wherein the mismatch standard comprises that k groups of similarity mean values arranged based on time sequence are sequentially reduced, and/or at least one group of similarity mean values in the k groups of similarity mean values are smaller than a preset similarity mean value.
Based on the apparatus provided in the foregoing embodiment, optionally, the obtaining module 71 is configured to:
acquiring N historical load related parameters of the thermal power generating unit before a first historical time period;
determining extrema of the N historical load-related parameters;
equally dividing the N historical load related parameters into m groups according to the extreme values of the historical load related parameters, wherein m is an integer greater than 1;
determining distribution density of each group of historical load related parameters
Wherein k is i Is the total number of samples in the ith group, i is an integer, and i is more than or equal to 1 and less than or equal to m;
screening the historical load related parameters according to the distribution density of the historical load related parameters to obtain screened historical load related parameters meeting a preset density standard, wherein the preset density standard comprises that the difference value of the distribution density of the historical load related parameters of each group is smaller than a preset distribution density difference value;
wherein the prediction module 72 is configured to:
and inputting part of the screened historical load related parameters into a parameter model matched with the thermal power generating unit to obtain predicted load related parameters of the thermal power generating unit in the first historical time period, which are predicted by the parameter model.
Based on the apparatus provided in the foregoing embodiment, optionally, the prediction module 72 is configured to:
obtaining sample load parameters from the filtered historical load related parameters, wherein the sample load parameters comprise part of the historical load related parameters in each group of the historical load related parameters;
and inputting the sample load parameters into the parameter model to obtain the predicted load related parameters of the thermal power generating unit predicted by the parameter model in the first historical time period.
Based on the apparatus provided in the foregoing embodiment, optionally, the training module 73 is configured to:
generating sample input parameters according to the historical load related parameters;
generating sample output parameters corresponding to the historical load related parameters one by one according to the predicted load related parameters and the actual load related parameters;
and training the parameter model according to the sample input parameters and the sample output parameters.
Through the device provided by the embodiment of the application, historical load related parameters, predicted output parameters and historical output parameters of the thermal power generating unit are obtained; training a parameter model according to the historical load related parameters, the prediction output parameters and the historical output parameters; monitoring real-time operation parameters and real-time output parameters of the thermal power generating unit; and inputting the real-time operation parameters into the trained parameter model, and determining the operation state of the thermal power generating unit according to the output parameters and the real-time output parameters of the trained parameter model. According to the scheme, parameters for training the parameter model are generated based on historical output parameters and prediction output parameters, the parameter model training process can have inheritance, the reliability of the parameter model is improved, the trained parameter model can predict the output parameters of the thermal power generating unit according to the real-time operation parameters, and the accuracy of the monitored operation state is improved by combining the real-time output parameters.
In order to solve the problem in the prior art, an embodiment of the present application further provides a monitoring system for an operating state of a thermal power generating unit, as shown in fig. 8, including:
a thermal power generating unit 81;
a thermal power unit operating condition monitoring device 82 as claimed in claim 7 communicatively connected to said thermal power unit.
By the system provided by the embodiment of the application, historical load related parameters, predicted output parameters and historical output parameters of the thermal power generating unit can be obtained; training a parameter model according to the historical load related parameters, the prediction output parameters and the historical output parameters; monitoring real-time operation parameters and real-time output parameters of the thermal power generating unit; and inputting the real-time operation parameters into the trained parameter model, and determining the operation state of the thermal power generating unit according to the output parameters and the real-time output parameters of the trained parameter model. The parameters used for training the parameter model are generated based on the historical output parameters and the prediction output parameters, the parameter model training process can have inheritance, the reliability of the parameter model is improved, the trained parameter model can predict the output parameters of the thermal power generating unit according to the real-time operation parameters, and the accuracy of the monitored operation state is improved by combining the real-time output parameters.
Preferably, an embodiment of the present invention further provides an electronic device, which includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor, where the computer program, when executed by the processor, implements each process of the above-mentioned method for monitoring an operating state of a thermal power generating unit, and can achieve the same technical effect, and in order to avoid repetition, details are not described here again.
The embodiment of the invention also provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the computer program implements each process of the monitoring method for the operating state of the thermal power generating unit, and can achieve the same technical effect, and is not described here again to avoid repetition. The computer-readable storage medium may be a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a component of' 8230; \8230;" does not exclude the presence of another like element in a process, method, article, or apparatus that comprises the element.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.
While the present invention has been described with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, which are illustrative and not restrictive, and it will be apparent to those skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (9)
1. A method for monitoring the operating state of a thermal power generating unit is characterized by comprising the following steps:
acquiring historical load related parameters of a thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period, wherein the load related parameters comprise power consumption parameters of the thermal power generating unit and/or equipment operation physical parameters of the thermal power generating unit;
inputting part of the historical load related parameters into a parameter model matched with the thermal power generating unit to obtain predicted load related parameters of the thermal power generating unit in the first historical time period, which are predicted by the parameter model;
taking the historical load related parameters as sample input parameters, and taking the predicted load related parameters and the actual load related parameters as sample output parameters to train the parameter model;
monitoring load related parameters of the thermal power generating unit, inputting the load related parameters into a trained parameter model, and determining the operating state of the thermal power generating unit according to the output result of the parameter model;
before acquiring historical load related parameters of a thermal power generating unit before a first historical period and actual load related parameters of the thermal power generating unit in the first historical period, the method further comprises the following steps:
acquiring actual load related parameters of the thermal power generating unit in a second historical time period and predicted load related parameters of the thermal power generating unit in the second historical time period, wherein the predicted load related parameters are predicted by the parameter model;
judging whether the parameter model is mismatched according to the similarity of the actual load related parameters in the second historical period and the predicted load related parameters in the second historical period;
the method for acquiring historical load related parameters of a thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period comprises the following steps:
when the parameter models are mismatched, acquiring historical load related parameters of the thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period.
2. The method according to claim 1, wherein the obtaining of the actual load-related parameters of the thermal power generating unit in a second historical period and the predicted load-related parameters of the thermal power generating unit in the second historical period predicted by the parameter model comprises:
acquiring an actual load parameter set X of the thermal power generating unit in a second historical period i ＝[x 1 ,x 2 ，…,x n ]And a model prediction parameter set
Wherein n is a positive integer greater than 1, and the actual load parameter set X i And the model prediction parameter set
Each item in (1) corresponds to each other one by one;
wherein, judging whether the parameter model is mismatched according to the similarity between the actual load related parameters in the second historical period and the predicted load related parameters in the second historical period comprises:
respectively determining the actual load parameter set X i And the corresponding set of model prediction parameters
Similarity of each item in (1)
Comparing the similarity based on time sequence
Dividing the image into k groups and determining similarity mean value A of each group vgk Wherein k is a positive integer less than or equal to n;
and determining the parameter model mismatch when the similarity mean values meet a mismatch standard, wherein the mismatch standard comprises that k groups of similarity mean values arranged based on a time sequence are sequentially reduced, and/or at least one group of similarity mean values in the k groups of similarity mean values are smaller than a preset similarity mean value.
3. The method of claim 1, wherein obtaining historical load-related parameters of a thermal power generating unit before a first historical period and actual load-related parameters of the thermal power generating unit during the first historical period comprises:
acquiring N historical load related parameters of the thermal power generating unit before a first historical time period;
determining extrema of the N historical load-related parameters;
equally dividing the N historical load related parameters into m groups according to the extreme values of the historical load related parameters, wherein m is an integer greater than 1;
determining distribution density of each group of historical load related parameters
Wherein k is i Is the total number of samples in the ith group, i is an integer, and i is more than or equal to 1 and less than or equal to m;
screening the historical load related parameters according to the distribution density of the historical load related parameters to obtain screened historical load related parameters meeting a preset density standard, wherein the preset density standard comprises that the difference value of the distribution density of the historical load related parameters of each group is smaller than a preset distribution density difference value;
inputting part of the historical load related parameters into a parameter model matched with the thermal power generating unit to obtain the predicted load related parameters of the thermal power generating unit predicted by the parameter model in the first historical time period, wherein the method comprises the following steps:
and inputting part of the screened historical load related parameters into a parameter model matched with the thermal power generating unit to obtain predicted load related parameters of the thermal power generating unit in the first historical time period, which are predicted by the parameter model.
4. The method according to claim 3, wherein inputting a portion of the filtered historical load-related parameters into a matched parameter model of the thermal power unit to obtain predicted load-related parameters of the thermal power unit predicted by the parameter model within the first historical period comprises:
obtaining sample load parameters from the filtered historical load related parameters, wherein the sample load parameters comprise part of the historical load related parameters in each group of the historical load related parameters;
and inputting the sample load parameters into the parameter model to obtain the predicted load related parameters of the thermal power generating unit predicted by the parameter model in the first historical time period.
5. The method of claim 4, wherein training the parametric model using the historical load-related parameters as sample input parameters and the predicted load-related parameters and the actual load-related parameters as sample output parameters comprises:
generating sample input parameters according to the historical load related parameters;
generating sample output parameters corresponding to the historical load related parameters one by one according to the predicted load related parameters and the actual load related parameters;
and training the parameter model according to the sample input parameters and the sample output parameters.
6. A monitoring device for the operating state of a thermal power generating unit is characterized by comprising:
the system comprises an acquisition module, a processing module and a control module, wherein the acquisition module is used for acquiring historical load related parameters of a thermal power unit before a first historical time period and actual load related parameters of the thermal power unit in the first historical time period, and the load related parameters comprise power consumption parameters of the thermal power unit and/or equipment operation physical parameters of the thermal power unit;
the prediction module is used for inputting part of the historical load related parameters into a parameter model matched with the thermal power generating unit so as to obtain the predicted load related parameters of the thermal power generating unit in the first historical time period, which are predicted by the parameter model;
the training module is used for taking the historical load related parameters as sample input parameters and taking the predicted load related parameters and the actual load related parameters as sample output parameters to train the parameter model;
the monitoring module is used for monitoring load related parameters of the thermal power generating unit, inputting the load related parameters into the trained parameter model, and determining the running state of the thermal power generating unit according to the output result of the parameter model;
the judging module is used for acquiring actual load related parameters of the thermal power unit in a second historical period and predicted load related parameters of the thermal power unit in the second historical period predicted by the parameter model before acquiring historical load related parameters of the thermal power unit before a first historical period and actual load related parameters of the thermal power unit in the first historical period;
judging whether the parameter model is mismatched according to the similarity of the actual load related parameters in the second historical period and the predicted load related parameters in the second historical period;
wherein the acquisition module is configured to: when the parameter models are mismatched, acquiring historical load related parameters of the thermal power generating unit before a first historical time period and actual load related parameters of the thermal power generating unit in the first historical time period.
7. A monitoring system for the operating state of a thermal power generating unit is characterized by comprising:
a thermal power generating unit;
a monitoring device for an operating condition of a thermal power generating unit according to claim 6, communicatively connected to the thermal power generating unit.
8. An electronic device, comprising: memory, processor and computer program stored on the memory and executable on the processor, which computer program, when executed by the processor, carries out the steps of the method according to any one of claims 1 to 5.
9. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, which computer program, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110155558.7A CN112909936B (en) | 2021-02-04 | 2021-02-04 | Method, device and system for monitoring running state of thermal power generating unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110155558.7A CN112909936B (en) | 2021-02-04 | 2021-02-04 | Method, device and system for monitoring running state of thermal power generating unit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112909936A CN112909936A (en) | 2021-06-04 |
| CN112909936B true CN112909936B (en) | 2022-10-04 |
Family
ID=76122362
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110155558.7A Active CN112909936B (en) | 2021-02-04 | 2021-02-04 | Method, device and system for monitoring running state of thermal power generating unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112909936B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115145211B (en) * | 2022-09-06 | 2022-12-09 | 上海桁壹高科技有限公司 | Control method, system, medium and electronic device of non-switching value device |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105095614A (en) * | 2014-04-18 | 2015-11-25 | 国际商业机器公司 | Method and device for updating prediction model |
| JP6625839B2 (en) * | 2015-07-08 | 2019-12-25 | 株式会社東芝 | Load actual data determination device, load prediction device, actual load data determination method, and load prediction method |
| CN106067077A (en) * | 2016-06-01 | 2016-11-02 | 新奥泛能网络科技股份有限公司 | A kind of load forecasting method based on neutral net and device |
| CN108416695B (en) * | 2018-02-24 | 2020-07-07 | 合肥工业大学 | Power load probability density prediction method, system and medium based on deep learning |
| EP3968247A4 (en) * | 2019-06-21 | 2022-12-21 | Siemens Aktiengesellschaft | Power load data prediction method and device, and storage medium |
| CN110392515B (en) * | 2019-07-23 | 2021-06-01 | 上海朗绿建筑科技股份有限公司 | Historical data-based cold and heat source machine room energy-saving control method and system |
| CN110781947B (en) * | 2019-10-22 | 2023-04-07 | 北京交通大学 | Power load prediction model training and power load prediction method and device |
| CN112116144B (en) * | 2020-09-15 | 2021-04-13 | 山东科技大学 | Regional power distribution network short-term load prediction method |
-
2021
- 2021-02-04 CN CN202110155558.7A patent/CN112909936B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN112909936A (en) | 2021-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11042145B2 (en) | Automatic health indicator learning using reinforcement learning for predictive maintenance | |
| US11604934B2 (en) | Failure prediction using gradient-based sensor identification | |
| CN116433009A (en) | Abnormality monitoring method and device for power transformation equipment and storage medium | |
| US20150286950A1 (en) | Apparatus for forecasting water demand | |
| CN112909936B (en) | Method, device and system for monitoring running state of thermal power generating unit | |
| CN112766618B (en) | Abnormality prediction method and device | |
| CN112559305B (en) | Method, device and equipment for predicting performance of key module in uninterruptible power supply system | |
| CN114298175A (en) | Power equipment state monitoring and fault early warning method and system based on edge calculation | |
| CN110865924A (en) | Health degree diagnosis method and health diagnosis framework for internal server of power information system | |
| US20220245096A1 (en) | Method and device for facilitating storage of data from an industrial automation control system or power system | |
| CN113758652B (en) | Oil leakage detection method and device for converter transformer, computer equipment and storage medium | |
| CN113110961B (en) | Equipment abnormality detection method and device, computer equipment and readable storage medium | |
| CN112734141B (en) | Diversified load interval prediction method and device | |
| CN117236380A (en) | Power system fault prediction method, system, electronic equipment and medium | |
| CN116819352A (en) | Self-adaptive setting method, device and equipment for battery threshold value and storage medium | |
| CN116881661A (en) | Performance automatic analysis method and system based on low-voltage power capacitor | |
| CN111506636A (en) | System and method for analyzing residential electricity consumption behavior based on autoregressive and neighbor algorithm | |
| CN113570473A (en) | Equipment fault monitoring method and device, computer equipment and storage medium | |
| CN111382891B (en) | Short-term load prediction method and short-term load prediction device | |
| Mayadevi et al. | SCADA-based operator support system for power plant equipment fault forecasting | |
| Al-Mohamad et al. | Robust zonotopic set-membership approach for model-based prognosis: Application on linear parameter-varying systems | |
| CN117094478B (en) | Energy scheduling management method, device, equipment and storage medium | |
| CN114253237A (en) | Equipment alarm prediction method, device, terminal and medium | |
| CN110232482B (en) | Equipment management method and device based on neural network | |
| CN117113212A (en) | Lightning arrester defect identification method and device and electronic equipment |
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 |