Turbine main engine load prediction method and system based on integrated empirical mode decomposition, electronic equipment and storage medium

Abstract

The invention provides a steam turbine main engine load prediction method and system based on integrated empirical mode decomposition, electronic equipment and a storage medium. The method comprises the following steps: carrying out EMD (Empirical Mode Decomposition) on a steam turbine main engine load signal added with white noise to obtain an IMF value; inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result; inputting the high-frequency component of the IMF value into a deep learning network to obtain a high-frequency IMF prediction result; and carrying out superposition reconstruction on the low-frequency IMF prediction result and the high-frequency IMF prediction result to obtain a steam turbine main engine load prediction value. According to the scheme provided by the invention, the ensemble empirical mode decomposition (EMD) algorithm can be applied to the steam turbine main engine load signal with relatively strong noise and relatively weak regularity, the decomposed signal components are classified into high-frequency and low-frequency signals according to the frequency, and a proper prediction algorithm is selected for the signal components of different frequency bands; and the prediction error is reduced to the greatest extent to improve the prediction effect.

Description

Technical Field

[0001] The present invention belongs to the technical field of steam turbine main engine load prediction, and in particular relates to a steam turbine main engine load prediction method, system, electronic equipment and storage medium based on integrated empirical mode decomposition. Background Art

[0002] Currently, there are few studies on load prediction for ship power systems, and load prediction for steam turbine main engines is even rarer. However, during ship operation, accurate prediction of the load of the steam turbine main engine can not only achieve early tracking of the steam turbine main engine load, which helps to improve the load control performance of the steam turbine main engine, but also provide technical support for the energy scheduling of the ship power system. Therefore, load prediction of the steam turbine main engine can effectively optimize the performance of the integrated power control system and improve the stability and efficiency of the ship power system operation. However, the load signal of the steam turbine main engine has a large amount of noise signal components. If artificial intelligence algorithms are used to directly infer and mine data on such signals, good results are often not achieved.

[0003] The above technical problems need to be solved urgently. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention proposes a steam turbine main engine load prediction method, system, electronic equipment and storage medium based on integrated empirical mode decomposition to solve the above technical problems.

[0005] A first aspect of the present invention discloses a method for predicting steam turbine main engine load based on integrated empirical mode decomposition, the method comprising:

[0006] Step S1, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF value;

[0007] Step S2: inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result;

[0008] Step S3: inputting the high-frequency component of the IMF value into a deep learning network to obtain a high-frequency IMF prediction result;

[0009] Step S4: superimpose and reconstruct the low-frequency IMF prediction result and the high-frequency IMF prediction result to obtain a predicted value of the steam turbine main engine load.

[0010] According to the method of the first aspect of the present invention, in step S1, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF value includes:

[0011] Step S11, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain IMF components with frequencies from high to low;

[0012] Step S12, repeating step S11 multiple times to obtain multiple groups of IMF components with frequencies from high to low;

[0013] Step S13: averaging the IMF components of each group from high to low frequency to obtain an IMF value.

[0014] According to the method of the first aspect of the present invention, in step S2, the machine learning algorithm is random forest regression;

[0015] Inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result includes:

[0016] The bagging algorithm is used to randomly sample the low-frequency components of the IMF values ​​to form multiple sample data sets, and multiple corresponding regression trees are built based on the multiple sample data sets; each regression tree is integrated based on the mean square error minimization principle to obtain the prediction result of each regression tree; the prediction results of each regression tree are weighted and summed to obtain the low-frequency IMF prediction result.

[0017] According to the method of the first aspect of the present invention, in step S3, the deep learning network is a gated recurrent unit.

[0018] A second aspect of the present invention discloses a steam turbine main engine load prediction system based on integrated empirical mode decomposition, the system comprising:

[0019] The first processing module is configured to perform EMD decomposition on the steam turbine main engine load signal with white noise added to obtain an IMF value;

[0020] A second processing module is configured to input the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result;

[0021] A third processing module is configured to input the high-frequency component of the IMF value into a deep learning network to obtain a high-frequency IMF prediction result;

[0022] The fourth processing module is configured to superimpose and reconstruct the low-frequency IMF prediction result and the high-frequency IMF prediction result to obtain a predicted value of the steam turbine main engine load.

[0023] According to the system of the second aspect of the present invention, the first processing module is specifically configured to perform EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF value, including:

[0024] Perform EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF components with high to low frequencies;

[0025] Repeat the process of “performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain IMF components with high to low frequencies” multiple times to obtain multiple groups of IMF components with high to low frequencies.

[0026] The IMF value is obtained by averaging the IMF components of each group from high to low frequency.

[0027] According to the system of the second aspect of the present invention, the second processing module is specifically configured such that the machine learning algorithm is random forest regression;

[0028] Inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result includes:

[0029] The bagging algorithm is used to randomly sample the low-frequency components of the IMF values ​​to form multiple sample data sets, and multiple corresponding regression trees are built based on the multiple sample data sets; each regression tree is integrated based on the mean square error minimization principle to obtain the prediction result of each regression tree; the prediction results of each regression tree are weighted and summed to obtain the low-frequency IMF prediction result.

[0030] According to the system of the second aspect of the present invention, the third processing module is specifically configured so that the deep learning network is a gated recurrent unit.

[0031] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps of any one of the methods for predicting steam turbine main engine load based on integrated empirical mode decomposition according to the first aspect of the present disclosure.

[0032] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any one of the methods for predicting steam turbine main engine load based on integrated empirical mode decomposition according to the first aspect of the present disclosure.

[0033] In summary, the solution proposed in the present invention can apply the ensemble empirical mode decomposition algorithm EEMD, which has excellent performance and stable performance in the field of signal decomposition, to the steam turbine main engine load signal with strong noise and weak regularity. The decomposed signal components are classified into high-frequency and low-frequency signals according to the frequency, and appropriate prediction algorithms are selected for signal components in different frequency bands to minimize the prediction error and improve the prediction effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 Flowchart of a method for predicting steam turbine main engine load based on integrated empirical mode decomposition according to an embodiment of the present invention;

[0036] Figure 2 A structural block diagram of a random forest regression (RFR) algorithm according to an embodiment of the present invention;

[0037] Figure 3 2. A structural block diagram of a gated recurrent unit (GRU) algorithm according to an embodiment of the present invention;

[0038] Figure 4 Schematic diagram showing comparison results between a load forecasting method based on integrated empirical mode decomposition and a single forecasting algorithm according to an embodiment of the present invention;

[0039] Figure 5 2 is a structural diagram of a steam turbine main engine load prediction system based on integrated empirical mode decomposition according to an embodiment of the present invention;

[0040] Figure 6 FIG. 4 is a structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0042] The first aspect of the present invention discloses a method for predicting steam turbine main engine load based on integrated empirical mode decomposition. Figure 1 FIG. 1 is a flow chart of a method for predicting steam turbine main engine load based on integrated empirical mode decomposition according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0043] Step S1, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF value;

[0044] Step S2: inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result;

[0045] Step S3: inputting the high-frequency component of the IMF value into a deep learning network to obtain a high-frequency IMF prediction result;

[0046] Step S4: superimpose and reconstruct the low-frequency IMF prediction result and the high-frequency IMF prediction result to obtain a predicted value of the steam turbine main engine load.

[0047] In step S1, the steam turbine main engine load signal with white noise added is subjected to EMD decomposition to obtain an IMF value.

[0048] In some embodiments, in step S1, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain an IMF value includes:

[0049] Step S11, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain IMF components with frequencies from high to low;

[0050] Step S12, repeating step S11 multiple times to obtain multiple groups of IMF components with frequencies from high to low;

[0051] Step S13: Since the mean value of the white noise spectrum is 0, the mean value of each group of IMF components from high to low frequency is calculated to obtain the IMF value.

[0052] Specifically, step S11 is repeated N times, where N is 300.

[0053] In step S2, the low-frequency component of the IMF value is input into a machine learning algorithm to obtain a low-frequency IMF prediction result.

[0054] In some embodiments, in step S2, the machine learning algorithm is random forest regression; considering the requirement of real-time prediction, a machine learning prediction algorithm with stronger real-time performance is selected, and random forest regression RFR is used for low-frequency components with strong periodicity;

[0055] like Figure 2 As shown, the inputting the low-frequency component of the IMF value into the machine learning algorithm to obtain the low-frequency IMF prediction result includes:

[0056] The bagging algorithm is used to randomly sample the low-frequency components of the IMF values ​​to form multiple sample data sets, and multiple corresponding regression trees are built based on the multiple sample data sets; each regression tree is integrated based on the mean square error minimization principle to obtain the prediction result of each regression tree; the prediction results of each regression tree are weighted and summed to obtain the low-frequency IMF prediction result.

[0057] In step S3, the high-frequency component of the IMF value is input into a deep learning network to obtain a high-frequency IMF prediction result.

[0058] In some embodiments, in step S3, the deep learning network is a gated recurrent unit. Considering the requirement of prediction accuracy, the deep learning network-gated recurrent unit GRU is selected for high-frequency components with weak periodicity and large noise.

[0059] Specifically, GRU is developed from LSTM. LSTM uses three gate functions: input gate, forget gate, and output gate to control input value, forget value, and output value. GRU network is simplified compared to LSTM. GRU network consists of two gate functions: update gate and reset gate. t It is the update gate, which is used to determine how much information in the previous hidden layer state is passed to the current hidden state h t in; r t To reset the gate, it is necessary to determine how much information of the hidden layer state at the previous moment needs to be forgotten. The algorithm structure of GRU is as follows: Figure 3 shown.

[0060] GRU first calculates the output value of each neuron through forward propagation, and then calculates the inference bias of each neuron through reverse calculation. In this process, the adaptive gradient optimization algorithm Adam is used to continuously update the weight of each neuron to optimize the model performance.

[0061] In order to verify the advantages of the proposed algorithm (ERG) in the prediction of steam turbine main engine load, three groups of mainstream single prediction algorithms were used for comparative prediction, and a total of 700 data were predicted under various working conditions. The single prediction algorithms selected were random forest algorithm RFR, recurrent gate unit algorithm GRU and recurrent neural network RNN. The prediction results were compared. Figure 4 As shown in the figure. The mean square error (RMSE) is the expected value of the square of the difference between the predicted value and the true value. The smaller the RMSE, the better the prediction performance. R2, also known as the coefficient of determination, is a metric used to measure the degree of fit between the predicted value and the true value. The larger the R2, the better the fit. As can be seen, the RMSE of our algorithm ERG is less than 1, while the RMSEs of the other algorithms are all above 2. The R2 of our algorithm ERG is close to 1, while the R2 of the other algorithms is less than 0.5, indicating that ERG has a strong ability to restore data details. In summary, ERG's prediction accuracy and fitting ability are superior to those of other algorithms, confirming the effectiveness of the ensemble empirical mode decomposition algorithm.

[0062] In summary, the solution proposed in the present invention can apply the ensemble empirical mode decomposition algorithm EEMD, which has excellent performance and stable performance in the field of signal decomposition, to the steam turbine main engine load signal with strong noise and weak regularity. The decomposed signal components are classified into high-frequency and low-frequency signals according to the frequency, and appropriate prediction algorithms are selected for signal components in different frequency bands to minimize the prediction error and improve the prediction effect.

[0063] The second aspect of the present invention discloses a steam turbine main engine load prediction system based on integrated empirical mode decomposition. Figure 5 FIG. 1 is a structural diagram of a steam turbine main engine load prediction system based on integrated empirical mode decomposition according to an embodiment of the present invention; FIG. Figure 5 As shown, the system 100 includes:

[0064] The first processing module 101 is configured to perform EMD decomposition on the steam turbine main engine load signal with white noise added thereto to obtain an IMF value;

[0065] The second processing module 102 is configured to input the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result;

[0066] The third processing module 103 is configured to input the high-frequency component of the IMF value into a deep learning network to obtain a high-frequency IMF prediction result;

[0067] The fourth processing module 104 is configured to superimpose and reconstruct the low-frequency IMF prediction result and the high-frequency IMF prediction result to obtain a predicted value of the steam turbine main engine load.

[0068] According to the system of the second aspect of the present invention, the first processing module 101 is specifically configured to perform EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF value, including:

[0069] Step S11, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain IMF components with frequencies from high to low;

[0070] Step S12, repeating step S11 multiple times to obtain multiple groups of IMF components with frequencies from high to low;

[0071] Step S13: Since the mean value of the white noise spectrum is 0, the mean value of each group of IMF components from high to low frequency is calculated to obtain the IMF value.

[0072] Specifically, step S11 is repeated N times, where N is 300.

[0073] According to the system of the second aspect of the present invention, the second processing module 102 is specifically configured as follows: the machine learning algorithm is random forest regression; considering the requirement of real-time prediction, a machine learning prediction algorithm with stronger real-time performance is selected, and random forest regression RFR is used for low-frequency components with strong periodicity;

[0074] like Figure 2 As shown, the inputting the low-frequency component of the IMF value into the machine learning algorithm to obtain the low-frequency IMF prediction result includes:

[0075] The bagging algorithm is used to randomly sample the low-frequency components of the IMF values ​​to form multiple sample data sets, and multiple corresponding regression trees are built based on the multiple sample data sets; each regression tree is integrated based on the mean square error minimization principle to obtain the prediction result of each regression tree; the prediction results of each regression tree are weighted and summed to obtain the low-frequency IMF prediction result.

[0076] According to the system of the second aspect of the present invention, the third processing module 103 is specifically configured such that the deep learning network is a gated recurrent unit. Considering the requirement of prediction accuracy, the deep learning network-gated recurrent unit GRU is selected for high-frequency components with weak periodicity and large noise.

[0077] Specifically, GRU is developed from LSTM. LSTM uses three gate functions: input gate, forget gate, and output gate to control input value, forget value, and output value. GRU network is simplified compared to LSTM. GRU network consists of two gate functions: update gate and reset gate. t It is the update gate, which is used to determine how much information in the previous hidden layer state is passed to the current hidden state h t in; r t To reset the gate, it is necessary to determine how much information of the hidden layer state at the previous moment needs to be forgotten. The algorithm structure of GRU is as follows: Figure 3 shown.

[0078] Example 1

[0079] The hardware includes: power board, prediction board, structural components, display and control panel, connecting cables and other plug-ins. The prediction board is the core part of the device, including CPU, memory and other communication units, and realizes load prediction and communication with the monitoring station.

[0080] The main engine load prediction system is installed on the power integrated monitoring table, and the communication between the main engine load prediction system and the main engine load monitoring service cabinet is completed through RS485 communication.

[0081] Algorithm model training. Because the GRU neural network in this algorithm requires a large amount of sample data for pre-training, historical host load data under various steady-state and variable conditions from the host load monitoring service cabinet is fed into the algorithm for pre-training. The neural network continuously adjusts neuron weights based on this massive amount of historical load data, ultimately achieving ideal model performance.

[0082] Dynamic update of model parameters. During ship operation, the real-time load data of the main engine is continuously transmitted to the algorithm model in the turbine main engine load prediction device, and the model parameters are dynamically updated through real-time training.

[0083] The process of steam turbine main engine load prediction and PID correction is as follows:

[0084] 1) Transmitting host load data to the load forecasting device for real-time forecasting;

[0085] 2) After the turbine main engine load forecast is completed, the forecast results are transmitted to the EHL control cabinet;

[0086] 3) After receiving the load forecast data, the host controller of the EHL control cabinet uses it as a feedforward signal to perform PID calculation, adjust the host speed in advance, and optimize the host control.

[0087] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps of any one of the methods for predicting steam turbine main engine load based on integrated empirical mode decomposition disclosed in the first aspect of the present invention.

[0088] Figure 6 FIG. 1 is a structural diagram of an electronic device according to an embodiment of the present invention. Figure 6 As shown, the electronic device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the electronic device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, near field communication (NFC) or other technologies. The display screen of the electronic device can be a liquid crystal display or an electronic ink display screen, and the input device of the electronic device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the electronic device housing, or an external keyboard, touchpad or mouse.

[0089] Those skilled in the art will understand that Figure 6The structure shown in the figure is only a structural diagram of the part related to the technical solution of the present disclosure, and does not constitute a limitation on the electronic device to which the solution of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0090] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any one of the methods for predicting steam turbine main engine load based on integrated empirical mode decomposition according to the first aspect of the present invention.

[0091] Please note that the technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification. The above embodiments only express several implementation methods of the present application. The description is relatively specific and detailed, but it cannot be understood as a limitation on the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, without departing from the concept of this application, several variations and improvements can be made, which all fall within the scope of protection of this application. Therefore, the scope of protection of the patent in this application shall be based on the attached claims.

[0092] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for predicting steam turbine main engine load based on integrated empirical mode decomposition, characterized in that: The method comprises: Step S1, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain an IMF value; Step S2, inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result; Step S3, inputting the high-frequency component of the IMF value into a deep learning network to obtain a high-frequency IMF prediction result; Step S4: superimpose and reconstruct the low-frequency IMF prediction result and the high-frequency IMF prediction result to obtain a predicted value of the steam turbine main engine load.

2. A method for predicting steam turbine main engine load based on integrated empirical mode decomposition according to claim 1, characterized in that: In the step S1, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF value includes: Step S11, performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain IMF components with frequencies from high to low; Step S12, repeating step S11 multiple times to obtain multiple groups of IMF components with frequencies from high to low; Step S13, averaging each group of IMF components from high to low frequencies to obtain an IMF value.

3. The method for predicting steam turbine main engine load based on integrated empirical mode decomposition according to claim 1, characterized in that: In step S2, the machine learning algorithm is random forest regression; Inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result comprises: The bagging algorithm is used to randomly sample the low-frequency components of the IMF values ​​to form multiple sample data sets, and multiple corresponding regression trees are built according to the multiple sample data sets; each regression tree is integrated based on the mean square error minimization principle to obtain the prediction result of each regression tree; the prediction results of each regression tree are weighted and summed to obtain the low-frequency IMF prediction result.

4. The method for predicting steam turbine main engine load based on integrated empirical mode decomposition according to claim 1, characterized in that: In step S3, the deep learning network is a gated recurrent unit.

5. A steam turbine main engine load prediction system based on integrated empirical mode decomposition, characterized in that: The system comprises: The first processing module is configured to perform EMD decomposition on the steam turbine main engine load signal with white noise added to obtain an IMF value; The second processing module is configured to input the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result; A third processing module is configured to input the high-frequency component of the IMF value into a deep learning network to obtain a high-frequency IMF prediction result; The fourth processing module is configured to superimpose and reconstruct the low-frequency IMF prediction result and the high-frequency IMF prediction result to obtain a prediction value of the steam turbine main engine load.

6. A steam turbine main engine load prediction system based on integrated empirical mode decomposition according to claim 5, characterized in that: The first processing module is specifically configured to perform EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF value, including: Perform EMD decomposition on the steam turbine main engine load signal with white noise added to obtain the IMF components from high to low frequencies; Repeat "performing EMD decomposition on the steam turbine main engine load signal with white noise added to obtain IMF components with high to low frequencies" multiple times to obtain multiple groups of IMF components with high to low frequencies; The IMF value is obtained by averaging the IMF components of each group from high to low frequency.

7. The steam turbine main engine load prediction system based on integrated empirical mode decomposition according to claim 5, characterized in that: The second processing module is specifically configured such that the machine learning algorithm is random forest regression; Inputting the low-frequency component of the IMF value into a machine learning algorithm to obtain a low-frequency IMF prediction result comprises: The bagging algorithm is used to randomly sample the low-frequency components of the IMF values ​​to form multiple sample data sets, and multiple corresponding regression trees are built according to the multiple sample data sets; each regression tree is integrated based on the mean square error minimization principle to obtain the prediction result of each regression tree; the prediction results of each regression tree are weighted and summed to obtain the low-frequency IMF prediction result.

8. The steam turbine main engine load prediction system based on integrated empirical mode decomposition according to claim 5, characterized in that: The third processing module is specifically configured such that the deep learning network is a gated recurrent unit.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the steps in a method for predicting steam turbine main engine load based on integrated empirical mode decomposition described in any one of claims 1 to 4 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in a method for predicting steam turbine main engine load based on integrated empirical mode decomposition described in any one of claims 1 to 4 are implemented.

Images