Steam turbine data acquisition method and device, equipment and storage medium

Abstract

The invention discloses a steam turbine data acquisition method and device, equipment and a storage medium. The method comprises the steps that historical sensor data are determined in historical operation record data of the steam turbine, and the historical sensor data at least comprise five kinds of sensor data; determining data with the highest sensitivity in the historical sensor data as target sensor data, and determining other sensor data in the historical sensor data as candidate sensor data; correlation coefficients and correlation coefficients of the target sensor data and all the candidate sensor data are determined, target candidate sensor data are determined according to all the correlation coefficients and all the correlation coefficients, and the target candidate sensor data at least comprise three kinds of sensor data; the target sensor data and the target candidate sensor data are determined as target data, and the target data are used for training a steam turbine data reconstruction model. According to the technical scheme, the accuracy of the reconstruction result of the steam turbine data reconstruction model can be improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of computer technology, and in particular to a method, device, equipment and storage medium for acquiring steam turbine data. Background Art

[0002] As a key power equipment that converts the thermal energy of steam into mechanical energy, steam turbine plays an irreplaceable role in many fields such as power generation and industrial drive. If a steam turbine fails and needs to be repaired, it will increase the operating costs of the enterprise, reduce the economic benefits of the enterprise, and in serious cases endanger the lives of the enterprise personnel. Therefore, in order to ensure the safe and stable operation of the steam turbine, increase the operating life of the steam turbine, and ensure the healthy operation of the steam turbine, it is of great significance to carry out fault warning for the steam turbine.

[0003] At present, when training the fault warning model of steam turbines, its data set requires both normal operation data and fault data. However, for newly used steam turbines, there is only normal data in the initial operation, and very little or no fault data, which makes it impossible to establish a fault warning model for new steam turbines. Therefore, in order to solve this problem, a steam turbine data reconstruction model with only normal data in the training data set is used to help realize fault warning.

[0004] However, although the currently used turbine data reconstruction model can reconstruct the turbine operation data to a certain extent, the accuracy of its reconstruction results still needs to be improved and cannot meet the growing accuracy requirements. Summary of the invention

[0005] The present invention provides a method, device, equipment and storage medium for obtaining steam turbine data, which can improve the accuracy of reconstruction results of a steam turbine data reconstruction model.

[0006] In a first aspect, an embodiment of the present invention provides a method for acquiring data of a steam turbine, comprising:

[0007] Determining historical sensor data in historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data;

[0008] Determine the data with the highest sensitivity in the historical sensor data as target sensor data, and determine other sensor data in the historical sensor data as candidate sensor data;

[0009] Determine a correlation coefficient and a relevance coefficient between the target sensor data and each of the candidate sensor data, and determine the target candidate sensor data according to each of the correlation coefficients and each of the relevance coefficients, wherein the target candidate sensor data includes at least three types of sensor data;

[0010] The target sensor data and the target candidate sensor data are determined as target data, wherein the target data is used for training a steam turbine data reconstruction model.

[0011] The embodiment of the present invention provides a method for acquiring data of a steam turbine, comprising: determining historical sensor data in the historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data; determining the data with the highest sensitivity in the historical sensor data as target sensor data, and determining other sensor data in the historical sensor data as candidate sensor data; determining the correlation coefficient and association coefficient between the target sensor data and each candidate sensor data, and determining the target candidate sensor data according to each correlation coefficient and each association coefficient, wherein the target candidate sensor data includes at least three types of sensor data; determining the target sensor data and the target candidate sensor data as target data, wherein the target data is used for training a steam turbine data reconstruction model. The above technical scheme first determines the historical sensor data in the historical operation record data of the steam turbine, providing a data basis for determining the target sensor data and the candidate sensor data later. Then, the data with the highest sensitivity in the historical sensor data is determined as the target sensor data, and other sensor data in the historical sensor data is determined as the candidate sensor data, which helps to improve the quality of the target data determined subsequently. Then, the correlation coefficient and the association coefficient between the target sensor data and each candidate sensor data are determined, and the target candidate sensor data is determined according to each correlation coefficient and each association coefficient, thereby avoiding the overfitting phenomenon of the steam turbine data reconstruction model during the training process due to the inclusion of too many types of sensor data in the target data. Finally, the target sensor data and the target candidate sensor data are determined as the target data, thereby improving the training efficiency and model accuracy of the steam turbine data reconstruction model. Therefore, the present invention solves the problem of insufficient accuracy of the existing steam turbine data reconstruction model.

[0012] In a second aspect, an embodiment of the present invention further provides a steam turbine data acquisition device, the device comprising:

[0013] An acquisition module, used for determining historical sensor data in the historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data;

[0014] A sensor determination module, configured to determine the data with the highest sensitivity in the historical sensor data as target sensor data, and to determine other sensor data in the historical sensor data as candidate sensor data;

[0015] A screening module, used to determine the correlation coefficient and the association coefficient between the target sensor data and each of the candidate sensor data, and determine the target candidate sensor data according to each of the correlation coefficients and each of the association coefficients, wherein the target candidate sensor data includes at least three types of sensor data;

[0016] The target data determination module is used to determine the target sensor data and the target candidate sensor data as target data, wherein the target data is used for training a steam turbine data reconstruction model.

[0017] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0018] at least one processor; and a memory communicatively coupled to the at least one processor;

[0019] Wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can implement any of the turbine data acquisition methods described in the first aspect.

[0020] In a fourth aspect, an embodiment of the present invention further provides a storage medium including computer executable instructions,

[0021] The computer executable instructions, when executed by a computer processor, implement any of the steam turbine data acquisition methods described in the first aspect.

[0022] It should be noted that the above-mentioned computer instructions may be stored in whole or in part on a computer-readable storage medium. The computer-readable storage medium may be packaged together with the processor of the steam turbine data acquisition device, or may be packaged separately from the processor of the steam turbine data acquisition device, and this application does not limit this.

[0023] The description of the second, third and fourth aspects in this application can refer to the detailed description of the first aspect; and the beneficial effects of the description of the second, third and fourth aspects can refer to the beneficial effect analysis of the first aspect, which will not be repeated here.

[0024] In this application, the name of the steam turbine data acquisition device does not limit the device or functional module itself. In actual implementation, these devices or functional modules may appear with other names. As long as the functions of each device or functional module are similar to those of this application, they fall within the scope of the claims of this application and their equivalent technologies.

[0025] These and other aspects of the present application will become more apparent from the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0027] Figure 1 A flow chart of a method for acquiring steam turbine data provided by an embodiment of the present invention;

[0028] Figure 2 A flow chart of another method for acquiring steam turbine data provided by an embodiment of the present invention;

[0029] Figure 3 A schematic structural diagram of a steam turbine data acquisition device provided by an embodiment of the present invention;

[0030] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0032] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0033] The terms "first" and "second" and the like in the specification and drawings of this application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0034] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0035] It should be mentioned before discussing exemplary embodiments in more detail that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe various operations (or steps) as sequential processes, many operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of various operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the accompanying drawings. The process can correspond to methods, functions, procedures, subroutines, subprograms, etc. In addition, the embodiments in the present invention and the features in the embodiments can be combined with each other without conflict.

[0036] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0037] In the description of the present application, unless otherwise specified, “plurality” means two or more.

[0038] Figure 1 A flow chart of a method for acquiring steam turbine data provided by an embodiment of the present invention is provided. This embodiment is applicable to situations where the accuracy of a steam turbine data reconstruction model needs to be improved. The method can be performed by a steam turbine data acquisition device and specifically includes the following steps:

[0039] Step 110: Determine historical sensor data in the historical operation record data of the steam turbine.

[0040] Among them, steam turbine refers to a rotary power mechanical equipment that converts the thermal energy of steam into mechanical energy. Historical operation record data refers to the collection of various relevant data recorded during the actual operation of the steam turbine in the past period of time. Historical sensor data refers to the data collected by various sensors installed on the steam turbine during the past period of operation, such as high exhaust pressure, high exhaust temperature, post-stage extraction temperature, unit load, main steam pressure, main steam temperature, main steam flow, reheat steam pressure, reheat steam temperature, regulating stage pressure and regulating stage temperature. Historical sensor data includes at least five types of sensor data.

[0041] Specifically, a database query statement can be used to obtain historical sensor data of a preset period from a database storing historical operation record data of the steam turbine. The preset period refers to a time interval determined according to actual conditions or needs, and is used to extract historical sensor data from the database storing historical operation records of the steam turbine.

[0042] For example, a structured query language can be used to obtain data from all sensors from October 1 to October 2, 2024 from a database storing historical operation record data of a steam turbine, and identify it as historical sensor data.

[0043] In this embodiment, by determining the historical sensor data, a data basis is provided for subsequently determining the target sensor data and the candidate sensor data.

[0044] Step 120: determine the data with the highest sensitivity in the historical sensor data as the target sensor data, and determine the other sensor data in the historical sensor data as the candidate sensor data.

[0045] Among them, sensitivity refers to the sensitivity of the sensor to changes in the measured object. The target sensor data refers to the sensor data with the highest sensitivity in the historical sensor data. The candidate sensor data refers to the sensor data other than the target sensor data in the historical sensor data.

[0046] Specifically, the data type of each acquired sensor data can be determined based on the historical sensor data, and then the configuration file of the corresponding sensor can be queried based on the aforementioned determined data type to obtain the sensitivity of each sensor. After that, these sensitivities are compared to find the sensor with the highest sensitivity, and then the data corresponding to the sensor with the highest sensitivity in the historical sensor data is determined as the target sensor data, and the other sensor data in the historical sensor data is determined as candidate sensor data.

[0047] In this embodiment, because sensor data with high sensitivity are more sensitive to slight changes in the internal state of the turbine, which means they can capture more details and features, thereby providing more representative and discriminative information, the data with the highest sensitivity in the historical sensor data is determined as the target sensor data, which can improve the quality of the subsequently determined target data.

[0048] Step 130 : Determine the correlation coefficient and the association coefficient between the target sensor data and each candidate sensor data, and determine the target candidate sensor data according to each correlation coefficient and each association coefficient.

[0049] Among them, the target candidate sensor data refers to the candidate sensor data that has a strong close connection with the target sensor data, which is screened and determined based on the correlation coefficient and the correlation coefficient between the target sensor data and each candidate sensor data, and the target candidate sensor data includes at least three types of sensor data. The correlation coefficient refers to a statistical indicator used to measure the degree of linear correlation between two variables. In this embodiment, the correlation coefficient is used to indicate the closeness of the linear correlation between the target sensor data and each candidate sensor data. The correlation coefficient refers to an indicator used to measure the closeness of other forms of association between the target sensor data and each candidate sensor data, except for the linear relationship.

[0050] Specifically, after determining the target sensor data and each candidate sensor data, the target sensor data and each candidate sensor data may be preprocessed, that is, the target sensor data and each candidate sensor data may be preprocessed by noise filtering, outlier processing, data standardization, etc., to improve data quality. Then, according to actual conditions or requirements, suitable correlation coefficient calculation methods and correlation coefficient calculation methods may be selected to determine the correlation coefficient and correlation coefficient between the target sensor data and each candidate sensor data. For example, the correlation coefficient calculation method may be cosine similarity, Pearson correlation coefficient, Spearman rank correlation coefficient or Kendall rank correlation coefficient, etc., and the correlation coefficient calculation method may be transfer entropy, mutual information or Granger causality. Then, the selected correlation coefficient calculation method and correlation coefficient calculation method are applied to determine the correlation coefficient and correlation coefficient between the target sensor data and each candidate sensor data.

[0051] Then, according to the actual situation or needs, a suitable comprehensive method can be selected to process the correlation coefficient and the relevance coefficient of the target sensor data and each candidate sensor data to obtain the comprehensive coefficient of the target sensor data and each candidate sensor data. For example, the comprehensive method can be a weighted average method, a logical AND operation, or a hierarchical analysis method. Then, the selected comprehensive method is applied to determine the comprehensive coefficient of the target sensor data and each candidate sensor data. After that, the candidate sensor data whose comprehensive coefficient is greater than the preset threshold is determined as the target candidate sensor data. Among them, the preset threshold is a threshold determined according to the actual situation or needs and used to screen the target candidate sensor data.

[0052] In this embodiment, by determining the correlation coefficient and the association coefficient between the target sensor data and each candidate sensor data, and determining the target candidate sensor data based on the correlation coefficients and the association coefficients, the overfitting phenomenon of the turbine data reconstruction model during the training process is avoided due to the inclusion of too many types of sensor data in the target data.

[0053] Step 140: Determine the target sensor data and the target candidate sensor data as target data.

[0054] Among them, the target data is used for training the turbine data reconstruction model.

[0055] Specifically, after obtaining the target sensor data and the target candidate sensor data, a suitable data merging method can be selected according to the actual situation or requirements to merge the target sensor data and the target candidate sensor data into a whole to obtain the target data. For example, the data merging method can be list concatenation or array connection. Then, the target data can be used as a training data set of the steam turbine data reconstruction model to train the steam turbine data reconstruction model.

[0056] In this embodiment, by determining the target data, a large amount of irrelevant or redundant data that may exist is discarded, thereby improving the training efficiency of the subsequent model. In addition, the target data is screened based on the analysis of the correlation coefficient and the correlation coefficient of each sensor data. These data have a strong internal connection and synergy. Therefore, using the target data to train the turbine data reconstruction model can more accurately capture the key features and laws in the turbine operation data, thereby improving the model's understanding and reconstruction capabilities of the turbine operation status, thereby improving the accuracy of the model.

[0057] The steam turbine data acquisition method provided by the embodiment of the present invention includes: determining historical sensor data in the historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data; determining the data with the highest sensitivity in the historical sensor data as the target sensor data, and determining other sensor data in the historical sensor data as candidate sensor data; determining the correlation coefficient and association coefficient between the target sensor data and each candidate sensor data, and determining the target candidate sensor data according to each correlation coefficient and each association coefficient, wherein the target candidate sensor data includes at least three types of sensor data; determining the target sensor data and the target candidate sensor data as the target data, wherein the target data is used for training the steam turbine data reconstruction model. The above technical scheme first determines the historical sensor data in the historical operation record data of the steam turbine, which provides a data basis for determining the target sensor data and the candidate sensor data later. Then, the data with the highest sensitivity in the historical sensor data is determined as the target sensor data, and other sensor data in the historical sensor data is determined as the candidate sensor data, which helps to improve the quality of the target data determined subsequently. Then, the correlation coefficient and the association coefficient between the target sensor data and each candidate sensor data are determined, and the target candidate sensor data is determined according to each correlation coefficient and each association coefficient, thereby avoiding the overfitting phenomenon of the steam turbine data reconstruction model during the training process due to the inclusion of too many types of sensor data in the target data. Finally, the target sensor data and the target candidate sensor data are determined as the target data, thereby improving the training efficiency and model accuracy of the steam turbine data reconstruction model. Therefore, the present invention solves the problem of insufficient accuracy of the existing steam turbine data reconstruction model.

[0058] Figure 2 A flowchart of another method for acquiring steam turbine data provided by an embodiment of the present invention is provided. This embodiment is specific based on the above embodiment. In this embodiment, the method may further include:

[0059] Step 210: Acquire initial sensor data of an initial period from historical operation record data.

[0060] The initial period refers to a period of time determined according to actual conditions or needs, and is used to select initial sensor data. The initial sensor data refers to the data of each sensor in the initial period of time.

[0061] Specifically, the initial sensor data of the initial period can be acquired from a database storing historical operation record data of the steam turbine using a database query statement.

[0062] In this embodiment, obtaining the initial sensor data provides a data basis for obtaining the historical sensor data later.

[0063] Step 211 : Determine the extreme value difference of the initial sensor data according to the maximum value and the minimum value of the initial sensor data.

[0064] Among them, the extreme value difference refers to the difference between the maximum and minimum values ​​in a set of data.

[0065] Specifically, the difference between the maximum value and the minimum value of each sensor data in the initial sensor data is calculated respectively to obtain the extreme value difference of each sensor data, that is, the extreme value difference of the initial sensor data.

[0066] Step 212: Determine whether the extreme value difference of the initial sensor data is within the difference variation range corresponding to the initial sensor data.

[0067] If yes, execute step 214; if no, execute step 213.

[0068] The difference variation range refers to the fluctuation range of the extreme value difference that is preset according to actual conditions or requirements.

[0069] Specifically, after obtaining the extreme value difference of the initial sensor data, it can be determined whether the extreme value difference of each sensor data in the initial sensor data is within its corresponding difference variation range. If so, the initial sensor data is determined to be historical sensor data, and if not, the initial period is updated.

[0070] In this embodiment, by determining whether the extreme value difference of the initial sensor data is within the difference variation range corresponding to the initial sensor data, a judgment basis is provided for subsequently determining whether the initial sensor data is historical sensor data.

[0071] Step 213: Update the initial time period.

[0072] Specifically, after determining that the extreme value difference of the initial sensor data is not within the difference variation range corresponding to the initial sensor data, the initial time period can be updated according to the preset time update method. For example, the preset time update method can be to add 1 hour to the initial time period, and the initial time period is from 11:00 on January 1, 2024 to 12:00 on January 1, 2024, then the updated initial time period is from 12:00 on January 1, 2024 to 13:00 on January 1, 2024.

[0073] After executing step 213 , the process returns to executing step 210 .

[0074] Step 214: Determine whether the initial sensor data is historical sensor data.

[0075] Specifically, after determining that the extreme value difference of the initial sensor data is within the difference variation range corresponding to the initial sensor data, it can be determined that the initial sensor data is historical sensor data, thereby improving the data quality of the historical sensor data.

[0076] Furthermore, after determining that the initial sensor data is historical sensor data, it also includes: determining the median and median absolute deviation of the historical sensor data; for current historical sensor data, calculating the difference between the current historical sensor data and the median to obtain the absolute deviation of the current historical sensor data; if the ratio of the absolute deviation of the current historical sensor data to the median absolute deviation is not less than an error threshold, updating the current historical sensor data based on the median.

[0077] The median refers to the value in the middle (if the number of data is an odd number) or the average of the two middle numbers (if the number of data is an even number). The absolute deviation refers to the absolute value of the difference between each data point and the median. The median absolute deviation refers to the median of the absolute values ​​of the difference between each data point and the median. The error threshold refers to the threshold set in advance according to actual conditions or needs, which is used to determine whether the current historical sensor data deviates from the normal range. For example, the error threshold of the historical sensor data can be four times the median absolute deviation of the historical sensor data.

[0078] Specifically, for each type of sensor data in the historical sensor data, the median and median absolute deviation are calculated respectively. Then, for the current historical sensor data in each type of sensor data, the difference between the current historical sensor data and the median is calculated to obtain the absolute deviation of the current historical sensor data, and then the absolute deviation of the current historical sensor data is calculated to be a ratio of the absolute deviation of the current historical sensor data to the corresponding median absolute deviation. If the ratio is not less than the error threshold, the current historical sensor data is updated based on the median.

[0079] In this embodiment, through the above steps, outliers in the historical sensor data can be removed, thereby improving the quality of the historical sensor data.

[0080] Exemplarily, historical sensor data is shown in Table 1.

[0081] Table 1 Historical sensor data

[0082]

[0083] According to the data in Table 1, it can be calculated that the median value of the temperature sensor data is 30 and the median absolute deviation is 2; the median value of the pressure sensor data is 1.4 and the median absolute deviation is 0.1; the median value of the vibration sensor data is 0.7 and the median absolute deviation is 0.1. Assume that the error threshold of the temperature sensor data is 1.5; the error threshold of the pressure sensor data is 0.8; and the error threshold of the vibration sensor data is 1.2. After updating the historical sensor data whose ratio of the absolute deviation in each historical sensor data to its corresponding median absolute deviation is not less than the error threshold based on the median, the historical sensor data obtained is shown in Table 2.

[0084] Table 2 Updated historical sensor data

[0085]

[0086] Step 215: determine the data with the highest sensitivity in the historical sensor data as the target sensor data, and determine the other sensor data in the historical sensor data as the candidate sensor data.

[0087] Step 216 : For the current candidate sensor data, calculate the cosine similarity between the target sensor data and the current candidate sensor data to obtain a correlation coefficient between the target sensor data and the current candidate sensor data.

[0088] Specifically, after the target sensor data and the candidate sensor data are determined, for the current candidate sensor data, the cosine similarity between the target sensor data and the current candidate sensor data can be calculated to obtain the correlation coefficient between the target sensor data and the current candidate sensor data. The specific calculation formula is as follows:

[0089]

[0090] Among them, A is the vector corresponding to the current candidate sensor data, B is the vector corresponding to the target sensor data, |A| is the modulus of the vector corresponding to the current candidate sensor data, |B| is the modulus of the vector corresponding to the target sensor data, and cos(A,B) is the correlation coefficient between the target sensor data and the current candidate sensor data.

[0091] In this embodiment, the above steps provide a screening data basis for subsequently determining target candidate sensor data.

[0092] Step 217: Determine the correlation coefficient between the target sensor data and the current candidate sensor data according to the conditional probability of each target sensor data among all target sensor data, the conditional probability of each target sensor data among all current candidate sensor data and all target sensor data, and the joint probability of each target sensor data among all current candidate sensor data and all target sensor data.

[0093] Specifically, for the current candidate sensor data, the correlation coefficient between the target sensor data and the current candidate sensor data can be determined according to the conditional probability of each target sensor data among all target sensor data, the conditional probability of each target sensor data among all current candidate sensor data and all target sensor data, and the joint probability of each target sensor data among all current candidate sensor data and all target sensor data. The specific calculation formula is as follows:

[0094]

[0095] Among them, TE A→B is the correlation coefficient between the target sensor data and the current candidate sensor data, b t is the target sensor data at time t, a t is the current candidate sensor data at time t, P(b t+1 ,a t ,b t ) is the joint probability of the target sensor data among all current candidate sensor data and all target sensor data, P(b t+1 |a t ,b t ) is the conditional probability of the target sensor data among all current candidate sensor data and all target sensor data, P(b t+1 |b t ) is the conditional probability of the target sensor data among all target sensor data.

[0096] In this embodiment, the above steps provide a screening data basis for subsequently determining target candidate sensor data.

[0097] Step 218: Determine target candidate sensor data according to each correlation coefficient and each association coefficient.

[0098] In one implementation, step 218 may specifically include: determining the correlation contribution value of each candidate sensor data according to the correlation coefficient between each candidate sensor data and the target sensor data and the correlation coefficient weight corresponding to each candidate sensor data; determining the correlation contribution value of each candidate sensor data according to the correlation coefficient between each candidate sensor data and the target sensor data and the correlation coefficient weight corresponding to each candidate sensor data; calculating the sum of the correlation contribution value and the correlation contribution value of each candidate sensor data to obtain the importance of each candidate sensor data; determining the candidate sensor whose importance is greater than a preset importance threshold as the target candidate sensor data.

[0099] Among them, the correlation coefficient weight refers to the weight value of the correlation coefficient determined according to actual conditions and experience in order to evaluate the correlation contribution of the candidate sensor data. The correlation contribution value refers to the value obtained after considering the correlation coefficient between the candidate sensor data and the target sensor data and the correlation coefficient weight. The correlation coefficient weight refers to the weight value of the correlation coefficient determined according to actual conditions and experience in order to evaluate the correlation contribution of the candidate sensor data. The correlation contribution value refers to the value calculated based on the correlation coefficient between the candidate sensor data and the target sensor data and the correlation coefficient weight. Importance refers to a comprehensive indicator obtained by calculating the sum of the correlation contribution value and the correlation contribution value of the candidate sensor data. The preset importance threshold refers to the importance standard value set according to time conditions or needs for screening target candidate sensor data.

[0100] Specifically, after obtaining the correlation coefficient and the relevance coefficient of each candidate sensor data with the target sensor data, the correlation contribution value of each candidate sensor data can be determined according to the correlation coefficient of each candidate sensor data with the target sensor data and the correlation coefficient weight corresponding to each candidate sensor data, for example: the correlation contribution value of the current candidate sensor data = the correlation coefficient of the current candidate sensor data with the target sensor data × the correlation coefficient weight corresponding to the current candidate sensor data. And according to the correlation coefficient of each candidate sensor data with the target sensor data and the correlation coefficient weight corresponding to each candidate sensor data, the correlation contribution value of each candidate sensor data is determined, for example: the correlation contribution value of the current candidate sensor data = the correlation coefficient of the current candidate sensor data with the target sensor data × the correlation coefficient weight corresponding to the current candidate sensor data. Then, the sum of the correlation contribution value and the relevance contribution value of each candidate sensor data is calculated to obtain the importance of each candidate sensor data, for example: the importance of the current candidate sensor data = the correlation contribution value of the current candidate sensor data + the relevance contribution value of the current candidate sensor data. Finally, the candidate sensor whose importance is greater than the preset importance threshold is determined as the target candidate sensor data.

[0101] In this embodiment, through the above steps, the quality of the determined target candidate sensor data can be improved.

[0102] Step 219: Determine the target sensor data and the target candidate sensor data as target data.

[0103] Furthermore, after the target sensor data and the target candidate sensor data are determined as the target data, it also includes: inputting the target data as training data into the autoencoder, performing model training on the autoencoder, and calculating the loss function; optimizing the model based on the back propagation algorithm until the preset optimization conditions are met, and obtaining the turbine data reconstruction model.

[0104] Among them, the autoencoder refers to an unsupervised learning neural network model, which consists of two parts: an encoder and a decoder. The encoder compresses the input data into a low-dimensional encoding representation, and the decoder attempts to reconstruct the original input data from the encoding. For example, the structure of the autoencoder can be a two-layer convolutional network and a two-layer deconvolutional network, and standardization operations are performed between each layer. The loss function refers to a function used to measure the difference between the autoencoder reconstructed data and the original input data. For example, the loss function can be the mean square error or the mean absolute error. The preset optimization condition refers to the condition for stopping training that is pre-set during the model training process. For example, the preset optimization condition can be that the loss function value reaches a certain threshold or the number of training iterations reaches a certain number. The turbine data reconstruction model refers to a trained and optimized autoencoder model that can reconstruct the sensor data of the turbine.

[0105] Specifically, after determining the target data, the target data can be cleaned and preprocessed to ensure that they meet the input requirements of the autoencoder, and then the target data is used as training data to train the autoencoder model. Specifically, the target data is first input into the encoder in the autoencoder for feature extraction to obtain feature data, and then the feature data is input into the decoder in the autoencoder to reconstruct the original target data based on the feature data to obtain reconstructed data. The loss function can then be calculated based on the reconstructed data and the target data. The specific calculation formula is as follows:

[0106]

[0107] Where N represents the number of samples of target data, x i represents the i-th target data, represents the i-th reconstructed data, Represents x i and The square of the Euclidean distance.

[0108] Then, it is determined whether the preset optimization conditions are met (such as the loss function is less than the preset threshold). If not, the model is optimized based on the back propagation algorithm until the preset optimization conditions are met to obtain the turbine data reconstruction model. For example, the gradient descent method or Adam can be selected to determine hyperparameters such as the learning rate in order to optimize the model.

[0109] In this embodiment, through the above steps, a steam turbine data reconstruction model with better accuracy is obtained.

[0110] Furthermore, after obtaining the turbine data reconstruction model, it also includes: obtaining real-time sensor data of the sensor corresponding to the target sensor data at the current moment and real-time candidate sensor data of the sensor corresponding to the target candidate sensor data at the current moment, inputting the real-time sensor data and the real-time candidate sensor data into the turbine data reconstruction model, and obtaining reconstructed sensor data of the real-time sensor data and reconstructed candidate sensor data of the real-time candidate sensor data; when it is determined that the difference between the reconstructed sensor data and the real-time sensor data is not less than a first preset difference, and the difference between the reconstructed candidate sensor data and the real-time candidate sensor data is not less than a second preset difference, it is determined that a turbine fault has occurred.

[0111] Among them, real-time sensor data refers to the value of the sensor corresponding to the target sensor data at the current time point. Real-time candidate sensor data refers to the value of the sensor corresponding to the target candidate sensor data at the current time point. Reconstructed sensor data refers to the reconstruction result of the real-time sensor data output by the model after the real-time sensor data is input into the turbine data reconstruction model. Reconstructed candidate sensor data refers to the reconstruction result of the real-time candidate sensor data output by the model after the real-time candidate sensor data is input into the turbine data reconstruction model. The first preset difference value refers to a threshold value pre-set according to actual conditions or needs for judging whether there is a problem with the real-time sensor data. The second preset difference value refers to a threshold value pre-set according to actual conditions or needs for judging whether there is a problem with the real-time candidate sensor data.

[0112] Specifically, after obtaining the turbine data reconstruction model, the real-time sensor data of the sensor corresponding to the target sensor data at the current moment and the real-time candidate sensor data of the sensor corresponding to the target candidate sensor data at the current moment can be obtained through the collection system of the turbine operation data, and then the real-time sensor data and the real-time candidate sensor data are input into the turbine data reconstruction model to obtain the reconstructed sensor data of the real-time sensor data and the reconstructed candidate sensor data of the real-time candidate sensor data. Then, it is determined whether the difference between the reconstructed sensor data and the real-time sensor data is not less than the first preset difference, and whether the difference between the reconstructed candidate sensor data and the real-time candidate sensor data is not less than the second preset difference. If neither is less than, it is determined that the turbine has failed.

[0113] In this embodiment, through the above steps, the accuracy of determining whether a turbine failure occurs is effectively improved.

[0114] The steam turbine data acquisition method provided by the embodiment of the present invention first obtains the initial sensor data of the initial period from the historical operation record data, providing a data basis for obtaining the historical sensor data later. Then, according to the maximum and minimum values ​​of the initial sensor data, the extreme value difference of the initial sensor data is determined. Then, it is determined whether the extreme value difference of the initial sensor data is within the difference variation range corresponding to the initial sensor data, which provides a judgment basis for determining whether the initial sensor data is historical sensor data later. If not, the initial period is updated, and the step of obtaining the initial sensor data of the initial period from the historical operation record data is returned. If it is, the initial sensor data is determined to be historical sensor data, which improves the data quality of the historical sensor data. Then, the data with the highest sensitivity in the historical sensor data is determined as the target sensor data, and the other sensor data in the historical sensor data is determined as the candidate sensor data, which helps to improve the quality of the target data determined subsequently. After that, for the current candidate sensor data, the cosine similarity between the target sensor data and the current candidate sensor data is calculated to obtain the correlation coefficient between the target sensor data and the current candidate sensor data. According to the conditional probability of each target sensor data among all target sensor data, the conditional probability of each target sensor data among all current candidate sensor data and all target sensor data, and the joint probability of each target sensor data among all current candidate sensor data and all target sensor data, the correlation coefficient between the target sensor data and the current candidate sensor data is determined. The target candidate sensor data is determined according to each correlation coefficient and each correlation coefficient, thereby avoiding the overfitting phenomenon of the turbine data reconstruction model during the training process due to the inclusion of too many types of sensor data in the target data. Finally, the target sensor data and the target candidate sensor data are determined as the target data, thereby improving the training efficiency and model accuracy of the turbine data reconstruction model. Therefore, the present invention solves the problem of insufficient accuracy of the existing turbine data reconstruction model.

[0115] On the other hand, after determining that the initial sensor data is historical sensor data, by determining the median and median absolute deviation of the historical sensor data, for the current historical sensor data, the difference between the current historical sensor data and the median is calculated to obtain the absolute deviation of the current historical sensor data. If the ratio of the absolute deviation to the median absolute deviation is not less than the error threshold, the current historical sensor data is updated based on the median, which can improve the quality of the historical sensor data.

[0116] In addition, the target data is input as training data into the autoencoder, and the autoencoder is trained to obtain a more accurate steam turbine data reconstruction model. After obtaining the steam turbine data reconstruction model, the real-time sensor data and the real-time candidate sensor data can be input into the steam turbine data reconstruction model to obtain the reconstructed sensor data of the real-time sensor data and the reconstructed candidate sensor data of the real-time candidate sensor data, and when it is determined that the difference between the reconstructed sensor data and the real-time sensor data is not less than the first preset difference, and the difference between the reconstructed candidate sensor data and the real-time candidate sensor data is not less than the second preset difference, it is determined that the steam turbine has failed, which effectively improves the accuracy of determining whether the steam turbine has failed.

[0117] Figure 3 A schematic structural diagram of a steam turbine data acquisition device provided in an embodiment of the present invention. The device and the steam turbine data acquisition methods of the above-mentioned embodiments belong to the same inventive concept. For details not fully described in the embodiments of the steam turbine data acquisition device, reference can be made to the embodiments of the above-mentioned steam turbine data acquisition method.

[0118] like Figure 3 As shown, the device comprises:

[0119] An acquisition module 310 is used to determine historical sensor data in the historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data;

[0120] A sensor determination module 320, configured to determine the data with the highest sensitivity in the historical sensor data as target sensor data, and to determine other sensor data in the historical sensor data as candidate sensor data;

[0121] A screening module 330, configured to determine a correlation coefficient and a relevance coefficient between the target sensor data and each of the candidate sensor data, and determine the target candidate sensor data according to each of the correlation coefficients and each of the relevance coefficients, wherein the target candidate sensor data includes at least three types of sensor data;

[0122] The target data determination module 340 is used to determine the target sensor data and the target candidate sensor data as target data, wherein the target data is used for training a steam turbine data reconstruction model.

[0123] Based on the above embodiment, the acquisition module 310 is specifically used for:

[0124] Acquiring initial sensor data of an initial period from the historical operation record data;

[0125] Determining an extreme value difference of the initial sensor data according to a maximum value and a minimum value of the initial sensor data;

[0126] If the extreme value difference of the initial sensor data is within the difference variation range corresponding to the initial sensor data, determining that the initial sensor data is the historical sensor data;

[0127] Otherwise, the initial time period is updated, and the process returns to the step of obtaining the initial sensor data of the initial time period from the historical operation record data until all the sensor data of the time period corresponding to the historical operation record data are traversed.

[0128] Based on the above embodiment, determining the correlation coefficient and the association coefficient between the target sensor data and each of the candidate sensor data includes:

[0129] For the current candidate sensor data, calculating the cosine similarity between the target sensor data and the current candidate sensor data to obtain a correlation coefficient between the target sensor data and the current candidate sensor data;

[0130] The correlation coefficient between the target sensor data and the current candidate sensor data is determined according to the conditional probability of each target sensor data among all the target sensor data, the conditional probability of each target sensor data among all the current candidate sensor data and all the target sensor data, and the joint probability of each target sensor data among all the current candidate sensor data and all the target sensor data.

[0131] On the basis of the above embodiment, determining the target candidate sensor data according to each of the correlation coefficients and each of the association coefficients includes:

[0132] Determine a correlation contribution value of each candidate sensor data according to the correlation coefficient between each candidate sensor data and the target sensor data and a correlation coefficient weight corresponding to each candidate sensor data;

[0133] Determining a correlation contribution value of each candidate sensor data according to the correlation coefficient between each candidate sensor data and the target sensor data and a correlation coefficient weight corresponding to each candidate sensor data;

[0134] Calculate the sum of the correlation contribution value and the association contribution value of each candidate sensor data to obtain the importance of each candidate sensor data;

[0135] The candidate sensors whose importance is greater than a preset importance threshold are determined as target candidate sensor data.

[0136] Based on the above embodiment, the device further includes:

[0137] An exception handling module is used to determine the median and median absolute deviation of the historical sensor data after determining the historical sensor data in the historical operation record data of the steam turbine; for current historical sensor data, calculate the difference between the current historical sensor data and the median to obtain the absolute deviation of the current historical sensor data; if the ratio of the absolute deviation of the current historical sensor data to the median absolute deviation is not less than an error threshold, update the current historical sensor data based on the median.

[0138] Based on the above embodiment, the device further includes:

[0139] A training module is used to input the target sensor data and the target candidate sensor data as training data into an autoencoder after the target sensor data and the target candidate sensor data are determined as target data, perform model training on the autoencoder, and calculate a loss function; optimize the model based on a back propagation algorithm until preset optimization conditions are met, and obtain a turbine data reconstruction model.

[0140] Based on the above embodiment, the device further includes:

[0141] A fault prediction module is used to obtain the real-time sensor data of the sensor corresponding to the target sensor data at the current moment and the real-time candidate sensor data of the sensor corresponding to the target candidate sensor data at the current moment after obtaining the turbine data reconstruction model, input the real-time sensor data and the real-time candidate sensor data into the turbine data reconstruction model, and obtain the reconstructed sensor data of the real-time sensor data and the reconstructed candidate sensor data of the real-time candidate sensor data; when it is determined that the difference between the reconstructed sensor data and the real-time sensor data is not less than a first preset difference, and the difference between the reconstructed candidate sensor data and the real-time candidate sensor data is not less than a second preset difference, determine that the turbine has a fault.

[0142] The steam turbine data acquisition device provided in the embodiment of the present invention can execute the steam turbine data acquisition method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0143] It is worth noting that in the embodiment of the above-mentioned steam turbine data acquisition device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0144] Figure 4A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 4 A block diagram of an exemplary electronic device 4 suitable for implementing embodiments of the present invention is shown. Figure 4 The electronic device 4 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0145] like Figure 4 As shown, the electronic device 4 is in the form of a general purpose computing electronic device. The components of the electronic device 4 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0146] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0147] The electronic device 4 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 4, including volatile and non-volatile media, removable and non-removable media.

[0148] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 4 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 not shown, usually called a "hard drive"). Although Figure 4 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.

[0149] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 42 generally perform the functions and / or methods of the embodiments described herein.

[0150] The electronic device 4 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), one or more devices that enable a user to interact with the electronic device 4, and / or any device that enables the electronic device 4 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the electronic device 4 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 20. Figure 4 As shown, the network adapter 20 communicates with other modules of the electronic device 4 via the bus 18. It should be understood that although Figure 4 Not shown, other hardware and / or software modules may be used in conjunction with the electronic device 4, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0151] The processing unit 16 executes various functional applications and page displays by running the program stored in the system memory 28, for example, implementing the steam turbine data acquisition method provided in the embodiment of the present invention, the method comprising:

[0152] Determining historical sensor data in historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data;

[0153] Determine the data with the highest sensitivity in the historical sensor data as target sensor data, and determine other sensor data in the historical sensor data as candidate sensor data;

[0154] Determine a correlation coefficient and a relevance coefficient between the target sensor data and each of the candidate sensor data, and determine the target candidate sensor data according to each of the correlation coefficients and each of the relevance coefficients, wherein the target candidate sensor data includes at least three types of sensor data;

[0155] The target sensor data and the target candidate sensor data are determined as target data, wherein the target data is used for training a steam turbine data reconstruction model.

[0156] Of course, those skilled in the art can understand that the processor can also implement the technical solution of the steam turbine data acquisition method provided by any embodiment of the present invention.

[0157] An embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, a method for acquiring data of a steam turbine provided in an embodiment of the present invention is implemented. The method includes:

[0158] Determining historical sensor data in historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data;

[0159] Determine the data with the highest sensitivity in the historical sensor data as target sensor data, and determine other sensor data in the historical sensor data as candidate sensor data;

[0160] Determine a correlation coefficient and a relevance coefficient between the target sensor data and each of the candidate sensor data, and determine the target candidate sensor data according to each of the correlation coefficients and each of the relevance coefficients, wherein the target candidate sensor data includes at least three types of sensor data;

[0161] The target sensor data and the target candidate sensor data are determined as target data, wherein the target data is used for training a steam turbine data reconstruction model.

[0162] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0163] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0164] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0165] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0166] It should be understood by those skilled in the art that the modules or steps of the present invention described above can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, optionally, they can be implemented by a program code executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0167] In addition, the acquisition, storage, use, and processing of data in the technical solution of the present invention are in compliance with the relevant provisions of national laws and regulations.

[0168] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for acquiring steam turbine data, characterized in that: include: Determining historical sensor data in historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data; Determine the data with the highest sensitivity in the historical sensor data as target sensor data, and determine other sensor data in the historical sensor data as candidate sensor data; Determine a correlation coefficient and a relevance coefficient between the target sensor data and each of the candidate sensor data, and determine the target candidate sensor data according to each of the correlation coefficients and each of the relevance coefficients, wherein the target candidate sensor data includes at least three types of sensor data; The target sensor data and the target candidate sensor data are determined as target data, wherein the target data is used for training a steam turbine data reconstruction model.

2. The method for acquiring steam turbine data according to claim 1, characterized in that: The historical sensor data is determined in the historical operation record data of the steam turbine, including: Acquiring initial sensor data of an initial period from the historical operation record data; Determining an extreme value difference of the initial sensor data according to a maximum value and a minimum value of the initial sensor data; If the extreme value difference of the initial sensor data is within the difference variation range corresponding to the initial sensor data, determining that the initial sensor data is the historical sensor data; Otherwise, the initial time period is updated, and the process returns to the step of acquiring initial sensor data of the initial time period from the historical operation record data until all sensor data of the time period corresponding to the historical operation record data are traversed.

3. The method for acquiring steam turbine data according to claim 1, characterized in that: Determining the correlation coefficient and the association coefficient between the target sensor data and each of the candidate sensor data includes: For the current candidate sensor data, calculating the cosine similarity between the target sensor data and the current candidate sensor data to obtain a correlation coefficient between the target sensor data and the current candidate sensor data; The correlation coefficient between the target sensor data and the current candidate sensor data is determined according to the conditional probability of each target sensor data among all the target sensor data, the conditional probability of each target sensor data among all the current candidate sensor data and all the target sensor data, and the joint probability of each target sensor data among all the current candidate sensor data and all the target sensor data.

4. The method for acquiring steam turbine data according to claim 1, characterized in that: Determining target candidate sensor data according to each of the correlation coefficients and each of the association coefficients includes: Determine a correlation contribution value of each candidate sensor data according to the correlation coefficient between each candidate sensor data and the target sensor data and a correlation coefficient weight corresponding to each candidate sensor data; Determining a correlation contribution value of each candidate sensor data according to the correlation coefficient between each candidate sensor data and the target sensor data and a correlation coefficient weight corresponding to each candidate sensor data; Calculate the sum of the correlation contribution value and the association contribution value of each candidate sensor data to obtain the importance of each candidate sensor data; The candidate sensors whose importance is greater than a preset importance threshold are determined as target candidate sensor data.

5. The method for acquiring steam turbine data according to claim 1, characterized in that: After the historical sensor data is determined in the historical operation record data of the steam turbine, it also includes: determining a median and a median absolute deviation of the historical sensor data; For the current historical sensor data, calculating the difference between the current historical sensor data and the median value to obtain the absolute deviation of the current historical sensor data; If the ratio of the absolute deviation of the current historical sensor data to the median absolute deviation is not less than an error threshold, the current historical sensor data is updated based on the median.

6. The method for acquiring steam turbine data according to claim 1, characterized in that: After determining the target sensor data and the target candidate sensor data as target data, the method further includes: Input the target data into the autoencoder as training data, perform model training on the autoencoder, and calculate the loss function; The model is optimized based on the back propagation algorithm until the preset optimization conditions are met, and the turbine data reconstruction model is obtained.

7. The method for acquiring steam turbine data according to claim 6, characterized in that: After obtaining the steam turbine data reconstruction model, it also includes: Acquire the real-time sensor data of the sensor corresponding to the target sensor data at the current moment and the real-time candidate sensor data of the sensor corresponding to the target candidate sensor data at the current moment, input the real-time sensor data and the real-time candidate sensor data into the steam turbine data reconstruction model, and obtain the reconstructed sensor data of the real-time sensor data and the reconstructed candidate sensor data of the real-time candidate sensor data; When it is determined that the difference between the reconstructed sensor data and the real-time sensor data is not less than a first preset difference, and the difference between the reconstructed candidate sensor data and the real-time candidate sensor data is not less than a second preset difference, it is determined that the steam turbine fails.

8. A steam turbine data acquisition device, characterized in that: The device comprises: An acquisition module, used for determining historical sensor data in the historical operation record data of the steam turbine, wherein the historical sensor data includes at least five types of sensor data; A sensor determination module, configured to determine the data with the highest sensitivity in the historical sensor data as target sensor data, and to determine other sensor data in the historical sensor data as candidate sensor data; A screening module, used to determine the correlation coefficient and the association coefficient between the target sensor data and each of the candidate sensor data, and determine the target candidate sensor data according to each of the correlation coefficients and each of the association coefficients, wherein the target candidate sensor data includes at least three types of sensor data; The target data determination module is used to determine the target sensor data and the target candidate sensor data as target data, wherein the target data is used for training a steam turbine data reconstruction model.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the steam turbine data acquisition method described in any one of claims 1-7.

10. A storage medium containing computer executable instructions, characterized in that: The computer executable instructions are used to execute the steam turbine data acquisition method described in any one of claims 1-7 when executed by a computer processor.

Images