Gas generator excitation self-adaptive control method and system

Abstract

The invention discloses a gas generator excitation adaptive control method and system, and relates to the technical field of gas power generation, and the method comprises the steps: obtaining the dynamic operation data of a gas generator, building a first prediction model, carrying out the first prediction of the dynamic operation data through the first prediction model, and obtaining a first prediction result; obtaining static parameter data of the gas generator, establishing a second prediction model, and performing second prediction on the static parameter data through the second prediction model to obtain a second prediction result; and a final excitation control strategy is obtained by combining the first prediction result and the second prediction result, and excitation adaptive control of the gas generator is completed. According to the method, the precision, stability and response speed of excitation control are improved through a multi-layer and multi-model adaptive control strategy, so that the gas generator set can maintain efficient and reliable operation under various complex working conditions and load conditions, and a powerful guarantee is provided for the stability and economy of a power system.

Description

Technical Field

[0001] The present invention relates to the technical field of gas power generation, and in particular to a gas generator excitation adaptive control method and system. Background Art

[0002] With the optimization of energy structure and the promotion of clean energy, gas-fired generator sets play an increasingly important role in the power system. As an efficient and clean way of power generation, gas-fired generator sets have the advantages of fast start-up, flexible adjustment, and low pollution emissions. However, the operating characteristics of gas-fired generator sets are significantly different from those of traditional thermal power units, and their excitation control systems face more complex and severe challenges. Traditional excitation control methods are mainly based on PID controllers with fixed parameters, which are difficult to adapt to the dynamic characteristics of gas-fired generator sets under different operating conditions. With the intelligent development of power systems, advanced control methods based on data-driven and artificial intelligence have gradually attracted attention, but in practical applications, there are still problems such as insufficient model accuracy and poor real-time performance.

[0003] The existing gas generator excitation control technology has the following main deficiencies: First, the dynamic characteristics of gas generator sets are not adequately considered, making it difficult to accurately capture the impact of factors such as gas flow changes and ambient temperature fluctuations on the excitation system; second, the fusion of static parameters and dynamic operating data is not sufficient, resulting in the control strategy being difficult to achieve optimal adjustment under different operating conditions; third, the existing prediction models often ignore seasonal changes and long-term dependence on historical data, affecting the accuracy of the prediction results; and finally, the lack of an effective adaptive mechanism makes it difficult to automatically adjust the control strategy according to changes in the system state and external environment. These problems make it difficult for gas generator sets to achieve efficient and stable excitation control when facing a complex and changeable operating environment, affecting the dynamic performance and economy of the unit. Summary of the invention

[0004] In view of the problems of insufficient dynamic characteristic capture, insufficient fusion of static parameters and dynamic data, limited prediction model accuracy and insufficient adaptive capability in the existing gas generator excitation control technology, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: a gas generator excitation adaptive control method is proposed. The method realizes accurate prediction and adaptive control of the excitation current by establishing a dual prediction model of dynamic operation data and static parameter data, combining the GRU network, historical data influence, dynamic factor influence and seasonal changes.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for adaptive excitation control of a gas generator, comprising: obtaining dynamic operation data of the gas generator, establishing a first prediction model, performing a first prediction on the dynamic operation data through the first prediction model, and obtaining a first prediction result; obtaining static parameter data of the gas generator, establishing a second prediction model, performing a second prediction on the static parameter data through the second prediction model, and obtaining a second prediction result; combining the first prediction result and the second prediction result to obtain a final excitation control strategy, and completing adaptive excitation control of the gas generator.

[0007] As a preferred solution of the gas generator excitation adaptive control method of the present invention, the dynamic operation data includes real-time output power, rotor current, stator voltage, power factor, gas flow, ambient temperature, load change rate, and grid frequency; the static parameter data includes generator rated capacity, excitation system parameters, rotor inertia constant, and generator impedance parameters.

[0008] As a preferred solution of the gas generator excitation adaptive control method of the present invention, the first prediction model is based on the GRU network to make a preliminary prediction of the gas generator excitation current, and the first revised prediction value is obtained according to the influence of historical data and the second revised prediction value is obtained according to the influence of dynamic factors, and seasonal changes are added to obtain, specifically: based on the output of the GRU network, the excitation current of the gas generator is preliminarily predicted to obtain a preliminary excitation current prediction value Based on the preliminary excitation current prediction According to the influence of historical data, the first revised forecast value is obtained Based on the first revised prediction value According to the influence of dynamic factors, the second revised forecast value is obtained Based on the second revised prediction value Introducing the seasonal function, we get the first prediction result I f .

[0009] As a preferred solution of the gas generator excitation adaptive control method of the present invention, the preliminary excitation current prediction value It is expressed by the following formula:

[0010]

[0011] in, represents the preliminary predicted value of excitation current, h t represents the hidden state output of the GRU network at time t, W o represents the output weight matrix, b o Represents the output bias term; the first corrected prediction value It is expressed by the following formula:

[0012]

[0013] in, represents the first corrected prediction value, α represents the adjustment coefficient, K(·) represents the radial kernel function, x i Indicates the calculation of historical data points, x t represents the feature vector at the current moment, n represents the number of historical data points used to calculate the similarity; the second revised prediction value It is expressed by the following formula:

[0014]

[0015] in, represents the second revised predicted value, represents the first revised prediction value, P t represents the output power factor at time t, V t represents the stator voltage factor at time t, F t represents the grid frequency factor at time t, Q t represents the reactive power at time t, Q t-1 Represents the reactive power at time t-1; the first prediction result I f It is expressed by the following formula:

[0016]

[0017] Among them, I f represents the first prediction result, represents the second revised prediction value, β1 and β2 are the adjustment coefficients of the daily cycle and the annual cycle respectively, T d Indicates the length of a day cycle, T y Indicates the length of a one-year cycle.

[0018] As a preferred solution of the gas generator excitation adaptive control method of the present invention, the second prediction model is obtained by obtaining a preliminary excitation current estimate based on the rated capacity of the generator and the excitation system parameters, combined with the rotor inertia constant and the generator impedance parameters, specifically: obtaining a preliminary excitation current estimate I based on the rated capacity S of the generator and the excitation system parameters K f1 ; Based on the preliminary excitation current estimate I f1 and rotor inertia constant, the excitation current estimation I after considering the dynamic response is obtained f2 ; Based on the excitation current estimation I after considering the dynamic response f2 and the generator impedance parameters, and obtain the second prediction result I fbase .

[0019] As a preferred solution of the gas generator excitation adaptive control method of the present invention, the preliminary excitation current estimation If1 It is expressed by the following formula:

[0020]

[0021] Among them, I f1 represents the preliminary excitation current estimation, K represents the excitation system parameters, S represents the rated capacity of the generator, S ref represents the reference capacity of the generator; the excitation current estimation after considering the dynamic response I f2 It is expressed by the following formula:

[0022]

[0023] Among them, I f2 represents the excitation current estimation after considering the dynamic response, I f1 represents the preliminary excitation current estimate, H represents the rotor inertia constant, is the rotor angular velocity change rate; λ represents the dynamic response adjustment coefficient; the second prediction result I fbase It is expressed by the following formula:

[0024]

[0025] Among them, I fbase represents the second prediction result, X d represents the d-axis synchronous reactance, X′ d represents the d-axis transient reactance, X q represents the q-axis synchronous reactance, X q ′ represents the q-axis transient reactance.

[0026] As a preferred solution of the gas generator excitation adaptive control method of the present invention, wherein: combining the first prediction result and the second prediction result to obtain the final excitation control strategy includes fusing the first prediction result and the second prediction result using a dynamic fusion method based on error feedback:

[0027] I ffinal =I f +(I fbase -I f )·F

[0028]

[0029] Among them, I ffinal Indicates the final excitation current control value, I f Represents the first prediction result, I fbase represents the second prediction result, F represents the dynamic fusion factor, k represents the adjustment coefficient, which is used to control the fusion speed, E represents the historical prediction error, and E fbase represents the historical average error of the second forecast model, Ef Represents the historical average error of the first forecast model, ε is a positive number used to prevent the denominator from being zero;

[0030] By setting the final excitation current control value I ffinal Input excitation system to adjust the excitation current of the generator.

[0031] Another object of the present invention is to provide a gas generator excitation adaptive control system, including: a dynamic data prediction module, used to obtain the dynamic operation data of the gas generator, establish a first prediction model, perform a first prediction on the dynamic operation data through the first prediction model, and obtain a first prediction result; a static data prediction module, used to obtain the static parameter data of the gas generator, establish a second prediction model, perform a second prediction on the static parameter data through the second prediction model, and obtain a second prediction result; a result fusion module, used to combine the first prediction result and the second prediction result to obtain a final excitation control strategy to complete the excitation adaptive control of the gas generator.

[0032] A computer device includes a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the above gas generator excitation adaptive control method are implemented.

[0033] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the above-mentioned gas generator excitation adaptive control method are implemented.

[0034] Beneficial effects of the present invention: By innovatively combining dynamic operation data and static parameter data, a dual prediction model architecture is established. The method adopts deep learning technology based on the GRU network, and introduces the influence of historical data, dynamic factor adjustment and seasonal changes to achieve multi-dimensional and multi-time scale prediction of excitation current. By introducing physical model parameters and dynamic response adjustment, the interpretability and adaptability of the model are enhanced. Finally, a dynamic fusion method based on error feedback is adopted to achieve intelligent fusion of prediction results. This multi-level, multi-model adaptive control strategy significantly improves the accuracy, stability and response speed of excitation control, enabling gas-fired generator sets to maintain efficient and reliable operation under various complex operating conditions and load conditions, providing a strong guarantee for the stability and economy of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0036] Figure 1 It is a schematic diagram of a flow chart in the first embodiment of the present invention;

[0037] Figure 2 FIG. 1 is a schematic diagram of the structure of a computer device in a third embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0039] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0040] Example 1

[0041] Reference Figure 1 , as an embodiment of the present invention, provides a gas generator excitation adaptive control method.

[0042] First, in the gas generator excitation adaptive control method provided by the present application, the terminal may include one or two processors and a memory for storing data, wherein the processor may include but is not limited to a processing system such as a microprocessor MCU or a programmable logic device FPGA. The above-mentioned terminal may also include a transmission device and an input and output device for communication functions. It can be understood by ordinary technicians in this field that Figure 1 The structure shown is for illustration only and does not limit the structure of the above-mentioned terminal.

[0043] The memory can be used to store computer programs, such as the computer program corresponding to the gas generator excitation adaptive control method in this embodiment. The processor executes various functional applications and data processing by running the computer program stored in the memory, that is, implementing the above method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage systems, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0044] The transmission device is used to receive or send data via a network. The above network includes a wireless network provided by a communication provider of the terminal. In one example, the transmission device includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0045] like Figure 1 As shown, an embodiment of the present invention provides a method for adaptively controlling excitation of a gas generator, comprising the following steps:

[0046] S1: Acquire dynamic operation data of the gas generator, establish a first prediction model, perform a first prediction on the dynamic operation data through the first prediction model, and obtain a first prediction result.

[0047] Specifically, the dynamic operation data includes real-time output power, rotor current, stator voltage, power factor, gas flow, ambient temperature, load change rate, and grid frequency. Real-time output power can directly reflect the current operating status and load condition of the generator. The rotor current reflects the load and excitation condition of the generator and is closely related to the stability of the generator. The stator voltage reflects the grid voltage condition and affects the output and stability of the generator. The power factor represents the ratio of active power to apparent power and reflects the operating efficiency of the generator. The gas flow is directly related to the input energy of the generator and affects the output power. The ambient temperature affects the cooling effect and efficiency of the generator. The load change rate reflects the dynamic characteristics of the load. The grid frequency reflects the balance state of the power system and affects the synchronous operation of the generator. The above dynamic operation data are preprocessed as the input of the model.

[0048] Furthermore, the first prediction model is a preliminary prediction of the excitation current of the gas generator based on the GRU network. The first revised prediction value is obtained according to the influence of historical data and the second revised prediction value is obtained according to the influence of dynamic factors. The historical data is the past dynamic operation data of the gas generator, and seasonal changes are added to obtain the specific:

[0049] Step 1: Make a preliminary prediction of the excitation current of the gas generator based on the output of the GRU network to obtain a preliminary prediction value of the excitation current

[0050]

[0051] in, represents the preliminary predicted value of excitation current, h trepresents the hidden state output of the GRU network at time t, W o represents the output weight matrix, b o Represents the output bias term.

[0052] Step 2: Based on the preliminary excitation current prediction value According to the influence of historical data, the first revised forecast value is obtained

[0053]

[0054] in, represents the first corrected prediction value, α represents the adjustment coefficient, K(·) represents the radial kernel function, x i Indicates the calculation of historical data points, x t Represents the feature vector at the current moment, and n represents the number of historical data points used to calculate the similarity.

[0055] Step 3: Based on the first revised forecast value According to the influence of dynamic factors, the second revised forecast value is obtained

[0056]

[0057] in, represents the second revised predicted value, represents the first revised prediction value, P t represents the output power factor at time t, V t represents the stator voltage factor at time t, F t represents the grid frequency factor at time t, Q t represents the reactive power at time t, Q t-1 Represents the reactive power at time t-1.

[0058] Step 4: Based on the second revised forecast value Introducing the seasonal function, we get the first prediction result I f :

[0059]

[0060] Among them, I f represents the first prediction result, represents the second revised prediction value, β1 and β2 are the adjustment coefficients of the daily cycle and the annual cycle respectively, T d Indicates the length of a day cycle, T y Indicates the length of a one-year cycle.

[0061] It should be noted that the excitation current I fThe output of the first prediction model refers to the actual excitation current required by the gas generator. It is the most direct dynamic data and is usually measured in amperes (A). The real-time output power and stator voltage are included in the output power factor P t and stator voltage factor V t The rotor current is one of the important inputs of the GRU network, reflecting the instantaneous operating status of the generator. The power factor is measured by the reactive power Q t The gas flow rate affects the output power factor P t The load change rate is reflected in Q t -Q t-1 In the formula, it indicates the change amplitude of reactive power. The grid frequency is used to calculate the grid frequency factor F. t The real-time status of the entire power grid is reflected on the basis of

[0062] Furthermore, I f The range of is positive real numbers, I f The specific value of represents the ratio relative to the rated excitation current of the generator. This representation method allows an intuitive understanding of the degree of deviation of the current excitation level relative to the normal operating state.

[0063] Furthermore, the first prediction model adopts multiple prediction levels, from preliminary prediction to multiple corrections, gradually improving the prediction accuracy, integrating the GRU deep learning network and traditional statistical methods. The GRU network can capture complex nonlinear relationships; the similarity of historical data points is calculated by radial kernel function, and the influence of historical data is incorporated into the prediction, which improves the timing performance of the model; the model takes into account multiple dynamic factors such as output power, stator voltage, grid frequency, reactive power, etc., and improves the adaptability to real-time operating conditions; the introduction of seasonal functions of daily and annual cycles can capture the periodic changes of excitation current; the operating characteristics of the generator may be affected by seasonal changes. For example, the daily cycle can adapt to the daily variation model of power demand. The annual cycle can cope with the impact of seasonal temperature changes on generator efficiency, and can predict and adapt to changes in electricity demand in different seasons (such as heating in winter and cooling in summer), which helps to plan the maintenance and operation strategies of generators in the long term. By considering these two periodic changes, the model can more accurately predict the operating needs of generators on different time scales, enabling the control system to adjust its strategy in advance to cope with foreseeable periodic changes. Through multi-level corrections and consideration of multiple factors, the model has strong adaptive capabilities, can adapt to different operating conditions, has strong accuracy and adaptability, and can better predict the excitation current of gas generators.

[0064] S2: Obtain static parameter data of the gas generator, establish a second prediction model, perform a second prediction on the static parameter data through the second prediction model, and obtain a second prediction result.

[0065] Specifically, static parameter data include generator rated capacity, excitation system parameters, rotor inertia constant, and generator impedance parameters. Generator rated capacity reflects the basic output capacity of the generator and is an important reference for predicting long-term performance. Excitation system parameters affect the voltage regulation capability and dynamic performance of the generator. The rotor inertia constant reflects the generator's ability to respond to load changes and system stability. Generator impedance parameters affect the generator's electrical characteristics and interaction with the power grid.

[0066] The second prediction model is to obtain a preliminary excitation current estimate based on the generator rated capacity and excitation system parameters, combined with the rotor inertia constant and generator impedance parameters, specifically:

[0067] Step 1: Get the preliminary excitation current estimate I based on the generator rated capacity S and excitation system parameters K f1 :

[0068]

[0069] Among them, I f1 represents the preliminary excitation current estimation, K represents the excitation system parameters, S represents the rated capacity of the generator, S ref Indicates the reference capacity of the generator.

[0070] Step 2: Estimate I based on preliminary excitation current f1 and rotor inertia constant, the excitation current estimation I after considering the dynamic response is obtained f2 :

[0071]

[0072] Among them, I f2 represents the excitation current estimation after considering the dynamic response, I f1 represents the preliminary excitation current estimate, H represents the rotor inertia constant, is the rate of change of rotor angular velocity; λ is the dynamic response adjustment coefficient.

[0073] Step 3: Estimation of excitation current I based on dynamic response f2 and the generator impedance parameters, and obtain the second prediction result I fbase :

[0074]

[0075] Among them, I fbase represents the second prediction result, X drepresents the d-axis synchronous reactance, X′ d represents the d-axis transient reactance, X q represents the q-axis synchronous reactance, X q ′ represents the q-axis transient reactance.

[0076] Furthermore, the second prediction model comprehensively considers multiple static parameters such as generator rated capacity, excitation system parameters, rotor inertia constant and generator impedance parameters, thereby improving the comprehensiveness and accuracy of the prediction; in step two, the rotor inertia constant and rotor angular velocity change rate are introduced to consider the dynamic response characteristics of the generator; in step three, the synchronous reactance and transient reactance of the d-axis and q-axis are used to further optimize the prediction results, considering the electrical characteristics of the generator; three steps are used to gradually refine the prediction results, and different factors are considered in each step to make the prediction more accurate; the model structure is clear, and each step has a clear physical meaning, which is easy to understand and implement.

[0077] S3: Combining the first prediction result and the second prediction result to obtain a final excitation control strategy, and completing the excitation adaptive control of the gas generator.

[0078] Specifically, the first prediction result and the second prediction result are combined to obtain a final excitation control strategy, including fusing the first prediction result and the second prediction result using a dynamic fusion method based on error feedback:

[0079] I ffinal =I f +(I fbase -I f )·F

[0080]

[0081] Among them, I ffinal Indicates the final excitation current control value, I f Represents the first prediction result, I fbase represents the second prediction result, F represents the dynamic fusion factor, k represents the adjustment coefficient, which is used to control the fusion speed, E represents the historical prediction error, and E fbase represents the historical average error of the second forecast model, E f Represents the historical average error of the first forecasting model. ε is a positive number used to prevent the denominator from being zero.

[0082] Further, the final excitation current control value is used as an input of the excitation system to adjust the excitation current of the generator, thereby controlling the output characteristics of the generator.

[0083] Furthermore, the final excitation control strategy adopts a dynamic fusion method based on error feedback, combining the results of the first prediction model and the second prediction model to fully utilize the advantages of the two models; through the dynamic fusion factor F, the weights of the two prediction models can be automatically adjusted according to their historical performance, thereby improving the adaptability of the control strategy; the introduction of historical prediction errors enables the control strategy to learn from past predictions and continuously improve the prediction accuracy; by adjusting the coefficient k, the fusion speed can be controlled to provide flexibility for different application scenarios; a small positive number ε is added to the fusion formula to prevent the denominator from being zero, thereby improving the stability and robustness of the algorithm; by combining two different types of prediction models (a first prediction model based on dynamic data and a second prediction model based on static parameters), the control strategy can more comprehensively consider various influencing factors and improve the overall control effect; by combining multiple parameters, considering dynamic characteristics, and using a dynamic fusion strategy, a comprehensive, adaptive, and precise gas generator excitation control method is provided.

[0084] Example 2

[0085] Reference Figure 1-2 , which is an embodiment of the present invention, provides a gas generator excitation adaptive control system, comprising:

[0086] A dynamic data prediction module, used to obtain dynamic operation data of the gas generator, establish a first prediction model, perform a first prediction on the dynamic operation data through the first prediction model, and obtain a first prediction result;

[0087] A static data prediction module, used to obtain static parameter data of the gas generator, establish a second prediction model, perform a second prediction on the static parameter data through the second prediction model, and obtain a second prediction result;

[0088] The result fusion module is used to combine the first prediction result and the second prediction result to obtain a final excitation control strategy to complete the excitation adaptive control of the gas generator.

[0089] For the specific definition of the gas generator excitation adaptive control system, please refer to the definition of the gas generator excitation adaptive control method mentioned above, which will not be repeated here. Each module in the above gas generator excitation adaptive control system can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0090] Example 3

[0091] Reference Figure 2, which is the third embodiment of the present invention. Based on the first two embodiments, the embodiment of the present invention provides a computer device, which may be a server, and its internal structure diagram may be as shown in FIG. Figure 2 As shown. The computer device includes a processor, a memory and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium.

[0092] The database of the computer device is used to store action detection data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the steps in any of the above-mentioned sparse tensor operation acceleration method embodiments are implemented.

[0093] Those skilled in the art will understand that Figure 2 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0094] In one embodiment, an embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in any of the above-mentioned sparse tensor operation acceleration method embodiments are implemented.

[0095] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0096] Example 4

[0097] As an embodiment of the present invention, a gas generator excitation adaptive control method is provided. In order to verify the beneficial effects of the present invention, a scientific demonstration is carried out through simulation experiments.

[0098] Specifically, in order to verify the effectiveness of the proposed gas generator excitation adaptive control method, the present invention conducted a 6-month experiment on a 300MW gas-steam combined cycle generator set of a certain power company. The experimental process is as follows:

[0099] First, the present invention collects the dynamic operation data and static parameter data of the generator set. The dynamic operation data includes real-time output power, rotor current, stator voltage, power factor, gas flow, ambient temperature, load change rate and grid frequency, which are collected every 5 seconds by the monitoring system of the generator set. The static parameter data includes the rated capacity of the generator (300MW), excitation system parameters (K=0.05), rotor inertia constant (H=6.5s) and generator impedance parameters (X d =1.8,X′ d =0.3,X q =1.7,X q ′=0.55).

[0100] Next, the present invention constructs a first prediction model and a second prediction model according to the method described in the content of the invention. For the first prediction model, the present invention uses a two-layer GRU network containing 128 hidden units, the input is the dynamic operation data of the past 24 hours, and the output is the excitation current prediction for the next 1 hour. The present invention uses the Adam optimizer, the learning rate is set to 0.001, and pre-training is performed on the first month's data. In step two, the present invention selects the Gaussian kernel as the radial kernel function, and n is set to 100. In step three, the present invention determines the optimal value through grid search. In step four, the present invention determines the adjustment coefficients of the daily cycle and the annual cycle through Fourier analysis.

[0101] For the second prediction model, the present invention performs preliminary excitation current estimation, dynamic response consideration and impedance parameter adjustment according to the steps in the invention content based on the static parameters of the generator. The generator reference capacity is set to 100MW, and the dynamic response adjustment coefficient λ is obtained by fitting historical data.

[0102] Finally, the present invention realizes a dynamic fusion method based on error feedback. The present invention initializes k=0.1, ε=1e-6, and dynamically adjusts these parameters during operation to obtain the best performance.

[0103] Furthermore, the proposed method is compared with the traditional PID control method and the single GRU prediction method. The experimental results are shown in the following table:

[0104] Table 1: Performance comparison of different excitation control methods

[0105]

[0106] It should be noted that Schemes 1-3 are incomplete versions of the present invention, corresponding to the method of using only the first prediction model, only the second prediction model, and only the simple average fusion method, respectively.

[0107] Further, it can be seen from the data in Table 1 that the method of the present invention is significantly superior to the traditional method and other comparative schemes in all indicators: the average error of the excitation current of the method of the present invention is only 0.98%, which is 69.8% lower than that of PID control and 55.0% lower than that of a single GRU prediction. This shows that the method can more accurately predict and control the excitation current, effectively reducing the over-excitation and under-excitation. The method of the present invention controls the voltage fluctuation within 0.005pu, which is 66.7% higher than that of PID control and 58.3% higher than that of a single GRU prediction. This means that the output voltage of the generator is more stable, which is conducive to the safe operation of the power grid. The method of the present invention increases the power factor to 0.99, which is close to the ideal value of 1, which is 7.6% higher than that of PID control and 5.3% higher than that of a single GRU prediction. This shows the advantages of the method in reactive power control, which helps to improve the power quality and transmission efficiency. The dynamic response time of the method of the present invention is shortened to 60ms, which is 60% faster than PID control and 50% faster than that of a single GRU prediction. This shows that the method can respond to system changes more quickly, which is conducive to maintaining system stability. The method of the present invention improves the energy utilization rate to 59.8%, which is 5.3% higher than PID control and 4.0% higher than single GRU prediction. This means that with the same fuel input, more electrical energy output can be generated, improving the power generation efficiency. The system stability index of the method of the present invention reaches 0.97, which is 29.3% higher than PID control and 18.3% higher than single GRU prediction. This shows that the method can better maintain the stable operation of the system and reduce the risk of oscillation and loss of step. The adaptability index of the method of the present invention is as high as 0.96, which is 41.2% higher than PID control and 21.5% higher than single GRU prediction. This shows that the method can better adapt to different operating conditions and load changes.

[0108] In summary, by comparing the results of Schemes 1-3, it can be seen that each component of the method of the present invention plays an important role. The first prediction model (Scheme 1) performs well in processing dynamic data, the second prediction model (Scheme 2) has advantages in considering static parameters, and the dynamic fusion strategy (compared to Scheme 3) further improves the overall performance; it fully demonstrates the innovation and superiority of the method of the present invention. By combining deep learning technology and traditional control theory, considering a variety of dynamic and static factors, and adopting an intelligent fusion strategy, this method achieves more accurate, faster and more stable gas generator excitation control. Not only does it improve the power generation efficiency and power quality, but it also enhances the system's anti-disturbance ability and adaptability, providing a strong guarantee for the efficient and stable operation of gas generators.

[0109] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A gas generator excitation adaptive control method, characterized in that: include: Acquire dynamic operation data of the gas generator, establish a first prediction model, perform a first prediction on the dynamic operation data by using the first prediction model, and obtain a first prediction result; Obtaining static parameter data of the gas generator, establishing a second prediction model, and performing a second prediction on the static parameter data using the second prediction model to obtain a second prediction result; The first prediction result and the second prediction result are combined to obtain a final excitation control strategy to complete the excitation adaptive control of the gas generator.

2. The gas generator excitation adaptive control method according to claim 1, characterized in that: The dynamic operation data include real-time output power, rotor current, stator voltage, power factor, gas flow, ambient temperature, load change rate, and grid frequency; the static parameter data include generator rated capacity, excitation system parameters, rotor inertia constant, and generator impedance parameters.

3. The gas generator excitation adaptive control method according to claim 2, characterized in that: The first prediction model is based on the preliminary prediction of the gas generator excitation current by the GRU network, and the first revised prediction value is obtained according to the influence of historical data and the second revised prediction value is obtained according to the influence of dynamic factors, and the first prediction result I is obtained by adding seasonal changes. f , specifically: Based on the output of the GRU network, a preliminary prediction of the excitation current of the gas generator is made to obtain a preliminary excitation current prediction value Based on the preliminary excitation current prediction According to the influence of the historical data, the first revised forecast value is obtained Based on the first revised prediction value According to the influence of dynamic factors, the second revised forecast value is obtained Based on the second revised prediction value Introducing the seasonal function, we get the first prediction result I f .

4. The gas generator excitation adaptive control method according to claim 3, characterized in that: The preliminary excitation current prediction value It is expressed by the following formula: in, represents the preliminary predicted value of the excitation current, h t represents the hidden state output of the GRU network at time t, W o represents the output weight matrix, b o represents the output bias term; The first corrected predicted value It is expressed by the following formula: in, represents the first corrected prediction value, α represents the adjustment coefficient, K(·) represents the radial kernel function, x i Indicates the calculation of historical data points, x t represents the feature vector at the current moment, and n represents the number of historical data points used to calculate the similarity; The second corrected predicted value It is expressed by the following formula: in, represents the second revised predicted value, represents the first revised predicted value; P t represents the output power factor at time t, V t represents the stator voltage factor at time t, F t represents the grid frequency factor at time t, Q t represents the reactive power at time t, Q t-1 represents the reactive power at time t-1; The first prediction result I f It is expressed by the following formula: Among them, I f represents the first prediction result, represents the second revised prediction value, β1 and β2 are the adjustment coefficients of the daily cycle and the annual cycle respectively, T d Indicates the length of a day cycle, T y Indicates the length of a one-year cycle.

5. The gas generator excitation adaptive control method according to claim 4, characterized in that: The second prediction model is obtained by obtaining a preliminary excitation current estimate based on the rated capacity of the generator and the excitation system parameters, combined with the rotor inertia constant and the generator impedance parameters, specifically: According to the rated capacity S of the generator and the excitation system parameter K, the preliminary excitation current estimate I is obtained. f1 ; Based on the preliminary excitation current estimation I f1 and rotor inertia constant, the excitation current estimation I after considering the dynamic response is obtained f2 ; Based on the excitation current estimation I after considering the dynamic response f2 and the generator impedance parameters, and obtain the second prediction result I fbase .

6. The gas generator excitation adaptive control method according to claim 5, characterized in that: The preliminary excitation current estimate I f1 It is expressed by the following formula: Among them, I f1 represents the preliminary excitation current estimation, K represents the excitation system parameters, S represents the rated capacity of the generator, S ref Indicates the reference capacity of the generator; The excitation current estimation I after considering the dynamic response f2 It is expressed by the following formula: Among them, I f2 represents the excitation current estimation after considering the dynamic response, I f1 represents the preliminary excitation current estimate, H represents the rotor inertia constant, is the rotor angular velocity change rate; λ represents the dynamic response adjustment coefficient; The second prediction result I fbase It is expressed by the following formula: Among them, I fbase represents the second prediction result, X d represents the d-axis synchronous reactance, X′ d represents the d-axis transient reactance, X q represents the q-axis synchronous reactance, X q ′ represents the q-axis transient reactance.

7. The gas generator excitation adaptive control method according to claim 6, characterized in that: Combining the first prediction result and the second prediction result to obtain a final excitation control strategy includes fusing the first prediction result and the second prediction result using a dynamic fusion method based on error feedback: I ffinal =I f +(I fbase -I f )·F Among them, I ffinal Indicates the final excitation current control value, I f Represents the first prediction result, I fbase represents the second prediction result, F represents the dynamic fusion factor, k represents the adjustment coefficient, which is used to control the fusion speed, E represents the historical prediction error, and E fbase represents the historical average error of the second forecast model, E f Represents the historical average error of the first forecast model, ε is a positive number used to prevent the denominator from being zero; By setting the final excitation current control value I ffinal Input excitation system to adjust the excitation current of the generator.

8. A system using the gas generator excitation adaptive control method according to any one of claims 1 to 7, characterized in that: include: A dynamic data prediction module, used for acquiring dynamic operation data of the gas generator, establishing a first prediction model, performing a first prediction on the dynamic operation data through the first prediction model, and obtaining a first prediction result; A static data prediction module, used for acquiring static parameter data of the gas generator, establishing a second prediction model, performing a second prediction on the static parameter data through the second prediction model, and obtaining a second prediction result; The result fusion module is used to combine the first prediction result and the second prediction result to obtain a final excitation control strategy to complete the excitation adaptive control of the gas generator.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the gas generator excitation adaptive control method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the gas generator excitation adaptive control method according to any one of claims 1 to 7 are implemented.

Images