A method and system for dynamically adjusting turbine speed based on machine learning

Abstract

The present invention relates to the technical field of turbine speed regulation, and discloses a method and system for dynamic regulation of turbine speed based on machine learning, the method comprising: obtaining environmental information and water flow information of a target water area, substituting the environmental information and water flow information into a pre-established water flow fluctuation model, and obtaining preset water flow fluctuation data of the target water area within a preset time period. Real-time water flow data of the target water area within a preset time period is obtained, and a comparison is made between the real-time water flow data and the preset water flow fluctuation data, and a preset turbine speed of the turbine is determined according to the comparison result. The real-time rotation speed of the turbine is obtained, and the turbine is regulated according to the speed difference between the real-time rotation speed and the preset turbine speed. The present invention combines environmental information and water flow data with a preset water flow fluctuation model, and compares the turbine speed and water flow conditions in real time, thereby realizing dynamic adjustment of the turbine speed and improving the stability of the generated voltage of the hydropower system under different water flow conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of turbine speed regulation, and in particular to a method and system for dynamically regulating turbine speed based on machine learning. Background Art

[0002] As a clean energy source, hydropower is widely adopted around the world. Its operation usually relies on stable water flow to ensure the continuous output of electricity. However, in some special water environments, such as seasonal rivers, tidal waters, or scenes with significant pressure fluctuations, the stability of water flow is often significantly disturbed by external factors. These unstable water flow conditions may make it difficult to maintain the turbine speed within the optimal range, thereby causing output voltage fluctuations, affecting the stability of the power grid and the quality of power supply.

[0003] Most existing hydropower systems are designed based on continuous and stable water flow, and their speed control relies on fixed parameter adjustment or simple feedback control mechanisms. When encountering large water flow fluctuations or sudden pressure changes, these systems are difficult to respond in time, which may cause the turbine speed to be too high or too low, further exacerbating the problem of voltage fluctuations. In addition, traditional methods have poor adaptability to nonlinear and complex water flow characteristics, and cannot achieve accurate control of real-time fluctuations, which restricts the operating efficiency of hydropower generation systems.

[0004] Therefore, under unstable water flow conditions, it is particularly important to develop a method that can sense water flow fluctuations in real time and intelligently adjust the turbine speed. Summary of the invention

[0005] In view of this, the present invention proposes a method and system for dynamic adjustment of turbine speed based on machine learning, aiming to solve the problem that the hydropower generation system in current technology is difficult to cope with the nonlinear characteristics caused by water flow fluctuations or pressure changes, and cannot achieve accurate real-time adjustment of turbine speed, thereby affecting the operating efficiency and voltage output stability of the hydropower generation system.

[0006] The present invention proposes a method for dynamically adjusting turbine speed based on machine learning, comprising:

[0007] Obtaining environmental information and water flow information of a target water area, and substituting the environmental information and water flow information into a pre-established water flow fluctuation model to obtain preset water flow fluctuation data of the target water area within a preset time period;

[0008] Acquire real-time water flow data of the target water area within the preset time period, and compare the real-time water flow data with the preset water flow fluctuation data, wherein:

[0009] If the real-time water flow data is consistent with the preset water flow fluctuation data, determining the preset speed corresponding to the preset water flow fluctuation data as the preset turbine speed;

[0010] If the real-time water flow data is inconsistent with the preset water flow fluctuation data, an adjustment coefficient is determined according to the data difference between the real-time water flow data and the preset water flow fluctuation data, the preset speed is adjusted according to the adjustment coefficient, and the adjusted preset speed is determined as the preset turbine speed;

[0011] The real-time rotation speed of the turbine is obtained, and the speed difference between the real-time rotation speed and the preset turbine speed is obtained, wherein:

[0012] If the real-time water flow data is consistent with the preset water flow fluctuation data, the speed adjustment parameter of the turbine within a preset time period is determined according to the preset water flow fluctuation data, and the turbine is controlled to adjust according to the speed adjustment parameter as an output instruction;

[0013] If the real-time water flow data is inconsistent with the preset water flow fluctuation data, the speed adjustment parameter of the turbine is determined according to the speed difference and the speed adjustment parameter is used as an output instruction to control the turbine for adjustment.

[0014] Furthermore, when the water flow fluctuation model is established in advance, it includes:

[0015] Acquire the environmental information, water flow information and turbine speed of each historical water flow fluctuation of the target water area, and establish the water flow fluctuation relationship according to the environmental information, water flow information and turbine speed of each historical water flow fluctuation;

[0016] Obtaining the linear relationship and distance measurement between each water flow fluctuation relational expression, and establishing the water flow fluctuation model according to the linear relationship and distance measurement between each water flow fluctuation relational expression;

[0017] The environmental information and water flow information of the target water area are substituted into the water flow fluctuation model to obtain preset water flow fluctuation data of the target water area within a preset time period and a preset rotation speed of the turbine.

[0018] Furthermore, when the water flow fluctuation model is established according to the linear relationship and distance measurement between the water flow fluctuation relationship equations, it includes:

[0019] Establishing a distance matrix according to the distance metric, and iteratively merging each of the water flow fluctuation relationship equations according to the distance matrix;

[0020] Obtaining each of the water flow fluctuation relationship equations after iterative merging, and determining it as a water flow fluctuation impact type;

[0021] Acquire environmental characteristics and water flow characteristics in the water flow fluctuation influence type, and classify each of the water flow fluctuation relationship equations based on the environmental characteristics and water flow characteristics;

[0022] The linear relationship between each of the water flow fluctuation relationship equations in the water flow fluctuation influence type is obtained, and the water flow fluctuation model is established according to the linear relationship between each of the water flow fluctuation influence types.

[0023] Furthermore, when establishing the water flow fluctuation model according to the linear relationship of each of the water flow fluctuation influence types, it includes:

[0024] Obtaining the environmental information and water flow information during any of the historical water flow fluctuations, and substituting the environmental information and water flow information during the historical water flow fluctuations into the water flow fluctuation model to obtain the preset water flow fluctuation data;

[0025] Comparing the water flow fluctuation data of the historical water flow fluctuation pattern with the preset water flow fluctuation data, wherein:

[0026] If the water flow fluctuation data is inconsistent with the preset water flow fluctuation data, the water flow fluctuation model is corrected.

[0027] Furthermore, if the real-time water flow data is inconsistent with the preset water flow fluctuation data, determining the adjustment coefficient according to the data difference between the real-time water flow data and the preset water flow fluctuation data includes:

[0028] Acquire the water flow velocity and water pressure in the real-time water flow data, and preset the water flow velocity and preset water pressure in the preset water flow fluctuation data;

[0029] Obtaining a flow velocity difference between the water body flow velocity and a preset water body flow velocity, and a pressure difference between the water body pressure and a preset water body pressure;

[0030] According to the relationship between the flow rate difference and the pre-configured preset flow rate difference range, and the pressure difference and the pre-configured preset pressure difference range, an acquisition index of the adjustment coefficient is determined, and the adjustment coefficient is determined according to the acquisition index, wherein:

[0031] If the flow rate difference is within the preset flow rate difference range, and the pressure difference is within the preset pressure difference range, then the acquisition index of the adjustment coefficient is determined to be zero, and the adjustment coefficient is 0;

[0032] If the flow velocity difference is within the preset flow velocity difference range, and the pressure difference is not within the pressure difference range, determining that the acquisition index of the adjustment coefficient is the pressure difference;

[0033] If the flow velocity difference is not within the preset flow velocity difference range, and the pressure difference is within the preset pressure difference range, then determining that the acquisition index of the adjustment coefficient is the flow velocity difference;

[0034] If the flow rate difference is not within the preset flow rate difference range, and the pressure difference is not within the preset pressure difference range, then the acquisition index of the adjustment coefficient is determined to be the flow rate difference and the pressure difference.

[0035] Furthermore, when determining the adjustment coefficient according to the acquisition index, it includes:

[0036] Substitute the acquisition index into the formula to obtain the adjustment coefficient, where the formula is as follows:

[0037] ;

[0038] Wherein, S is the adjustment coefficient, 1 is the weight coefficient, M is the acquisition index, and M min To obtain the minimum value of the indicator, M max To obtain the maximum value of the indicator.

[0039] Further, determining the speed adjustment parameter of the turbine within a preset time period according to the preset water flow fluctuation data, and controlling the turbine to adjust according to the speed adjustment parameter as an output instruction, includes:

[0040] Obtaining a preset water flow velocity in each preset time period in the preset water flow fluctuation data, and fitting a linear axis between the preset water flow velocities in each preset time period according to a least squares method;

[0041] Mapping the speed difference to the linear axis, obtaining the speed adjustment amount of each of the preset water flow speeds, and converting the speed adjustment amount into the adjustment parameter;

[0042] The turbine is controlled to be adjusted according to the speed adjustment parameter as an output instruction.

[0043] Furthermore, when determining the adjustment parameter, it includes:

[0044] Acquiring a real-time response speed of the turbine speed, and determining whether to modify the adjustment parameter according to a relationship between the real-time response speed and a preset response speed;

[0045] When the real-time response speed is equal to the preset response speed, determining not to modify the adjustment parameter;

[0046] When the real-time response speed is less than the preset response speed, a correction coefficient is determined according to a response speed difference between the real-time response speed and the preset response speed, and the adjustment parameter is corrected according to the correction coefficient.

[0047] Furthermore, when determining the correction coefficient according to the response speed difference between the real-time response speed and the preset response speed, it includes:

[0048] Determining the correction coefficient according to a relationship between the response speed difference and a preset first response speed difference and a preset second response speed difference;

[0049] When the response speed difference is less than or equal to the first preset response speed difference, the correction coefficient is determined to be Q1;

[0050] When the response speed difference is greater than the first preset response speed difference, and the response speed difference is less than or equal to the second preset response speed difference, the correction coefficient is determined to be Q2;

[0051] When the response speed difference is greater than the second preset response speed difference, the correction coefficient is determined to be Q3;

[0052] The first preset response speed difference is smaller than the second preset response speed difference, and 1<Q1<Q2<Q3<1.5.

[0053] Compared with the prior art, the beneficial effect of the present invention is that by acquiring the environmental information and water flow information of the target water area in real time and using the pre-established water flow fluctuation model, the turbine speed can be dynamically adjusted under different water flow fluctuation conditions, thereby ensuring that the power generation system can maintain high stability and efficiency under any water flow conditions. In addition, by comparing the real-time water flow data with the preset water flow fluctuation data in real time, the uncertainty of water flow fluctuation can be dealt with. If the real-time water flow data is consistent with the preset fluctuation data, it can be adjusted directly according to the preset speed, avoiding unnecessary calculation and energy waste; and when there is a difference in water flow fluctuation, the turbine speed will be automatically adjusted according to the difference, avoiding the problem of unstable turbine operation caused by excessive water flow fluctuation. This not only enhances the ability to respond quickly to water flow fluctuations, but also improves the flexibility and robustness of the hydropower generation system through an intelligent adjustment process. On the other hand, according to the data accumulated in long-term operation, the water flow fluctuation model is gradually optimized to make it more and more adaptable to different water areas. By mining the regularity in the data, the adjustment strategy is gradually adjusted during the process of water flow changes, avoiding blind and inefficient operations in traditional methods. This data-based adjustment method not only helps the hydropower system to effectively cope with complex water flow characteristics and achieve dynamic real-time adjustment, but also improves the accuracy of turbine operation and the overall efficiency of the power generation system. Moreover, through this intelligent dynamic adjustment method, the hydropower system can automatically adjust when facing different water flow fluctuations, avoiding problems such as over-adjustment or slow response, ensuring accurate control of turbine speed, and reducing equipment loss and energy waste.

[0054] On the other hand, the present application also provides a turbine speed dynamic adjustment system based on machine learning, comprising:

[0055] An acquisition module is configured to acquire environmental information and water flow information of a target water area, and substitute the environmental information and water flow information into a pre-established water flow fluctuation model to acquire preset water flow fluctuation data of the target water area within a preset time period;

[0056] a turbine control module, electrically connected to the acquisition module, the turbine control module being configured to acquire real-time water flow data of the target water area within the preset time period, and compare the real-time water flow data with the preset water flow fluctuation data, wherein: if the real-time water flow data is consistent with the preset water flow fluctuation data, the turbine control module determines the preset speed corresponding to the preset water flow fluctuation data as the preset turbine speed; if the real-time water flow data is inconsistent with the preset water flow fluctuation data, the turbine control module determines an adjustment coefficient according to a data difference between the real-time water flow data and the preset water flow fluctuation data, adjusts the preset speed according to the adjustment coefficient, and determines the adjusted preset speed as the preset turbine speed;

[0057] Wherein, the turbine control module is also configured to obtain the real-time rotation speed of the turbine, and obtain the speed difference between the real-time rotation speed and the preset turbine speed, wherein: if the real-time water flow data is consistent with the preset water flow fluctuation data, the turbine control module determines the speed adjustment parameter of the turbine within a preset time period according to the preset water flow fluctuation data, and controls the turbine to adjust according to the speed adjustment parameter as an output instruction; if the real-time water flow data is inconsistent with the preset water flow fluctuation data, the turbine control module determines the speed adjustment parameter of the turbine according to the speed difference, and controls the turbine to adjust according to the speed adjustment parameter as an output instruction.

[0058] It can be understood that the method and system for dynamic adjustment of turbine speed based on machine learning in the above-mentioned embodiments of the present invention have the same beneficial effects and will not be described in detail. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0060] Figure 1 A flowchart of a method for dynamically adjusting turbine speed based on machine learning provided in an embodiment of the present invention;

[0061] Figure 2 A functional block diagram of a turbine speed dynamic adjustment system based on machine learning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0063] like Figure 1 As shown, in some embodiments of the present application, this embodiment provides a method for dynamically adjusting turbine speed based on machine learning, including:

[0064] Step S100, obtaining environmental information and water flow information of the target water area, and substituting the environmental information and water flow information into a pre-established water flow fluctuation model to obtain preset water flow fluctuation data of the target water area within a preset time period.

[0065] Specifically, the pre-established water flow fluctuation model includes: obtaining the environmental information, water flow information and turbine speed of each historical water flow fluctuation of the target water area, and establishing a water flow fluctuation relationship based on the environmental information, water flow information and turbine speed of the historical water flow fluctuation. Obtaining the linear relationship and distance measurement between each water flow fluctuation relationship, and establishing a water flow fluctuation model based on the linear relationship and distance measurement between each water flow fluctuation relationship. Substituting the environmental information and water flow information of the target water area into the water flow fluctuation model, obtaining the preset water flow fluctuation data of the target water area within the preset time period and the preset turbine speed.

[0066] Specifically, when establishing a water flow fluctuation model based on the linear relationship and distance measurement between each water flow fluctuation relationship formula, it includes: establishing a distance matrix based on the distance measurement, and iteratively merging each water flow fluctuation relationship formula based on the distance matrix. Obtaining each water flow fluctuation relationship formula after iterative merging, and determining it as a water flow fluctuation impact type. Obtaining environmental characteristics and water flow characteristics in the water flow fluctuation impact type, and classifying each water flow fluctuation relationship formula based on the environmental characteristics and water flow characteristics. Obtaining the linear relationship between each water flow fluctuation relationship formula in the water flow fluctuation impact type, and establishing a water flow fluctuation model based on the linear relationship of each water flow fluctuation impact type.

[0067] It can be seen that by establishing a distance matrix and optimizing and integrating the historical water flow fluctuation relationship by iterative merging, this method can effectively extract the factors affecting water flow fluctuation and classify and determine different types of water flow fluctuation influence. Compared with the traditional model based on a single feature, this method fully considers the diversity of complex water flow fluctuations, laying a solid foundation for the speed prediction and regulation of hydropower turbines. Secondly, by taking environmental characteristics and water flow characteristics as the basis for classification, the main influencing factors of water flow fluctuations can be accurately captured, thereby achieving targeted model optimization. Especially under complex water flow conditions, the classified water flow fluctuation relationship can significantly improve the pertinence and accuracy of the model, avoiding the prediction bias problem caused by uneven feature distribution in traditional models. In addition, by substituting the real-time environmental information and water flow information of the target water area into the model, the water flow fluctuation data within the preset time period and the preset turbine speed can be quickly generated. This real-time and dynamic adaptability enables the model to provide reliable adjustment parameters in a timely manner in sudden water flow fluctuations or pressure changes, thereby effectively reducing the impact of turbine speed fluctuations on power generation efficiency and system stability. Finally, through the introduction of iterative optimization and linear relationship analysis, the model can fully tap the potential value of historical data during the construction process. By using the distance matrix constructed through distance measurement, the model can more intuitively measure the similarities between different water flow fluctuation relationships, thereby effectively retaining key features during the iterative merging process. This scientific data processing method not only improves the robustness of the model, but also provides highly reliable data support for the subsequent precise adjustment of turbine speed.

[0068] It can be understood that by using the historical data of the target water area, including environmental information, water flow information and the corresponding turbine speed, water flow fluctuation relational expressions are generated, which characterize the dynamic characteristics of water flow fluctuations and their impact on the turbine speed. Secondly, by deeply analyzing the water flow fluctuation relational expressions, a distance matrix is ​​generated to quantify the similarities and differences between the relational expressions. Through the iterative merging algorithm, the water flow fluctuation relational expressions with similar characteristics are clustered into impact types. In this process, not only the dimension reduction and structured processing of the data are achieved, but also the main driving factors and characteristic patterns of water flow fluctuations can be revealed, so that the model can better adapt to the complex water flow conditions in different water areas. At the same time, after the clustering is completed, the environmental characteristics and water flow characteristics of each type of water flow fluctuation impact type are further extracted and classified. The purpose of this step is to divide the complex water flow fluctuation characteristics into several interpretable and predictable types to ensure the operability and accuracy of the model. By extracting the linear relationship between various impact types, the overall water flow fluctuation model is finally established. This model can not only dynamically respond to changes in different environmental characteristics and water flow conditions, but also has strong generalization ability and can be applied to a variety of different water scenarios. Finally, the real-time environmental information and water flow information of the target water area are substituted into the water flow fluctuation model, which can quickly generate water flow fluctuation data and turbine preset speed within a preset period of time. This process is based on the model's deep learning of historical data and dynamic adaptation of the relationship, ensuring the accuracy and real-time nature of the output data. In this way, the model can provide precise guidance for turbine speed control, especially when water flow fluctuations are frequent or nonlinear characteristics are significant, greatly improving the hydropower system's responsiveness to water flow fluctuations and operating efficiency.

[0069] Specifically, the water flow fluctuation relationship is as follows:

[0070] q=(z1,z2,z3);where q is the water flow fluctuation relationship, z1 is the environmental information, z2 is the water flow information, and z3 is the turbine speed.

[0071] Specifically, when obtaining the distance measurement between each water flow fluctuation relationship formula, the distance measurement between each water flow fluctuation relationship formula can be obtained through Euclidean distance, Manhattan distance, cosine similarity, KL divergence (relative entropy), etc.

[0072] Specifically, according to the linear relationship of each water flow fluctuation influence type, when establishing the water flow fluctuation model, it includes: obtaining the environmental information and water flow information of any historical water flow fluctuation, and substituting the environmental information and water flow information of the historical water flow fluctuation into the water flow fluctuation model to obtain the preset water flow fluctuation data. The water flow fluctuation data of the historical water flow fluctuation type is compared with the preset water flow fluctuation data, wherein: if the water flow fluctuation data is inconsistent with the preset water flow fluctuation data, the water flow fluctuation model is corrected.

[0073] It is understandable that the water flow fluctuation model is established based on historical water flow fluctuation data, and preliminary preset water flow fluctuation data is generated by using these historical data as a basis. However, due to the dynamic and complex nature of the water flow environment, it is difficult to fully cover all possible water flow fluctuation situations by relying solely on historical data. To this end, the model adopts an adaptive correction mechanism: the environmental information and water flow information during historical water flow fluctuations are substituted into the preset data generated by the model, and compared with the actual historical data to detect the model's prediction deviation.

[0074] It can be seen that this comparison process is the key to dynamic correction. When the comparison results show that there is an inconsistency between the preset data generated by the model and the actual data, the correction process will be triggered. The correction process improves the model structure by adjusting the model parameters, revising the water flow fluctuation relationship, or optimizing the linear relationship. After the correction, the prediction ability of the water flow fluctuation model will be further improved, so as to more accurately reflect the actual water flow characteristics of the target water area. In this way, the water flow fluctuation model can be continuously iterated and optimized with the accumulation of historical data, forming the characteristics of self-learning and self-adaptation. In addition, this dynamic correction mechanism is particularly important for turbine speed regulation in complex water flow environments. In real application scenarios, water flow fluctuations may be affected by many factors, such as the terrain characteristics of the water area, seasonal changes, and sudden water flow disturbances. These factors cause water flow fluctuations to be highly nonlinear and random. If the water flow fluctuation model cannot be updated and corrected in real time, its prediction ability will drop rapidly, thereby affecting the speed control accuracy of the turbine. By constantly comparing historical data with preset data, the model can quickly discover deviations and adjust itself to ensure that the prediction results are highly consistent with the actual situation. This dynamic correction mechanism combines historical data with real-time data to form a dynamically updated model structure. This water flow fluctuation model can not only capture regular water flow changes, but also quickly adapt to abnormal water flow fluctuations, providing more precise guidance for turbine speed regulation, thereby avoiding the problem of abnormal operating speed of hydropower turbines due to poor prediction of water flow fluctuations.

[0075] Step S200: acquiring real-time water flow data of the target water area within a preset time period, and comparing the real-time water flow data with preset water flow fluctuation data.

[0076] Specifically, when comparing the real-time water flow data with the preset water flow fluctuation data, the method includes: if the real-time water flow data is consistent with the preset water flow fluctuation data, determining the preset speed corresponding to the preset water flow fluctuation data as the preset turbine speed. If the real-time water flow data is inconsistent with the preset water flow fluctuation data, determining an adjustment coefficient according to the data difference between the real-time water flow data and the preset water flow fluctuation data, adjusting the preset speed according to the adjustment coefficient, and determining the adjusted preset speed as the preset turbine speed.

[0077] Specifically, if the real-time water flow data is inconsistent with the preset water flow fluctuation data, then the adjustment coefficient is determined according to the data difference between the real-time water flow data and the preset water flow fluctuation data, including: obtaining the water body flow velocity and water body pressure in the real-time water flow data, and the preset water body flow velocity and preset water body pressure in the preset water flow fluctuation data. Obtain the flow velocity difference between the water body flow velocity and the preset water body flow velocity, and the pressure difference between the water body pressure and the preset water body pressure. According to the relationship between the flow velocity difference and the pre-configured preset flow velocity difference range, and the pressure difference and the pre-configured preset pressure difference range, determine the acquisition index of the adjustment coefficient, and determine the adjustment coefficient according to the acquisition index, wherein: if the flow velocity difference is within the preset flow velocity difference range, and the pressure difference is within the preset pressure difference range, then the acquisition index of the adjustment coefficient is determined to be zero, and the adjustment coefficient is 0. If the flow velocity difference is within the preset flow velocity difference range, and the pressure difference is not within the pressure difference range, then the acquisition index of the adjustment coefficient is determined to be the pressure difference. If the flow rate difference is not within the preset flow rate difference range, and the pressure difference is within the preset pressure difference range, the flow rate difference is determined as the acquisition index of the adjustment coefficient. If the flow rate difference is not within the preset flow rate difference range, and the pressure difference is not within the preset pressure difference range, the flow rate difference and the pressure difference are determined as the acquisition index of the adjustment coefficient.

[0078] Specifically, when determining the adjustment coefficient according to the acquisition index, it includes: substituting the acquisition index into the formula to obtain the adjustment coefficient, wherein the formula is as follows: Among them, S is the adjustment coefficient, 1 is the weight coefficient, M is the acquisition index, and M min To obtain the minimum value of the indicator, M max To obtain the maximum value of the indicator.

[0079] It can be seen that the real-time water flow data (including water flow velocity and pressure) of the target water area is obtained and compared with the preset water flow data (preset water flow velocity and pressure) predicted by the model to detect the difference between the two. If the two are consistent, the preset speed predicted by the model is directly used as the preset speed of the turbine; if they are inconsistent, the adjustment coefficient is dynamically calculated according to the data difference to correct the preset speed. In the process of determining the adjustment coefficient, the key influencing factors of the adjustment are distinguished by the threshold judgment of the preset range. The acquisition index of the adjustment coefficient can be the pressure difference, the flow velocity difference, or a combination of the two, depending on whether the difference exceeds the preset range. When the difference exceeds the range, it will be used as a basis, substituted into the adjustment formula, and the final adjustment coefficient will be calculated. The weight coefficient in the formula is used to balance the influence of each acquisition index on the adjustment, and the minimum and maximum values ​​ensure that the adjustment range is within a reasonable range.

[0080] It is understandable that by comparing the real-time water flow data with the preset water flow fluctuation data, it is possible to quickly determine whether the current water flow condition is consistent with the model prediction. This process can verify the accuracy of the preset model in real time and ensure that the parameters are adjusted according to the actual situation during operation. When the real-time data is consistent with the preset data, the preset speed is directly used to simplify the calculation process and avoid unnecessary complex adjustments. When the two are inconsistent, the data difference is used to further analyze the adjustment requirements, thereby providing an accurate basis for the subsequent dynamic adjustment of the turbine speed. Secondly, by introducing the flow velocity difference and the pressure difference, the multi-dimensional factors affecting the water flow fluctuation are fully considered, and a more refined adjustment mechanism is constructed. In the calculation process of the adjustment coefficient, the acquisition indicators are set for different difference ranges to ensure that the adjustment process has clear logic and reasonable step-by-step control. For example, when the flow velocity difference and the pressure difference are both within the set range, it is determined that no adjustment is required, which effectively avoids the interference of frequent adjustments on the stability of equipment operation; when the flow velocity difference or the pressure difference exceeds the preset range, it can be accurately corrected for the specific source of the difference, further improving the adaptability of the turbine operation. Again, the formula for determining the adjustment coefficient reflects the scientificity and flexibility of dynamic adjustment. By introducing the weight coefficient and obtaining the boundary range of the index, the formula effectively balances the amplitude of adjustment and the accuracy of control. In this process, the calculation of the adjustment coefficient not only takes into account the size of the current difference, but also avoids the negative impact of extreme adjustment on the operation of the turbine by setting the boundary range. In addition, by reasonably allocating the weight coefficient, the degree of influence of flow rate and pressure can be differentiated according to actual needs, so as to achieve a more personalized speed adjustment scheme. Finally, by adjusting the turbine speed in real time to always match the current water flow state, the energy loss during operation is minimized. At the same time, accurate difference analysis and dynamic acquisition of the adjustment coefficient also effectively reduce equipment wear and extend the service life of the turbine. In the overall application, this method can meet the requirements of efficient operation in complex water flow environments and is widely applicable to operation optimization scenarios of various types of hydropower generation or water flow driven equipment.

[0081] Step S300: obtaining the real-time rotation speed of the turbine, and obtaining the speed difference between the real-time rotation speed and the preset turbine speed.

[0082] Specifically, when obtaining the speed difference between the real-time rotation speed and the preset turbine speed, it includes: if the real-time water flow data is consistent with the preset water flow fluctuation data, the speed adjustment parameter of the turbine within the preset time period is determined according to the preset water flow fluctuation data, and the turbine is controlled to adjust according to the speed adjustment parameter as an output instruction. If the real-time water flow data is inconsistent with the preset water flow fluctuation data, the speed adjustment parameter of the turbine is determined according to the speed difference, and the turbine is controlled to adjust according to the speed adjustment parameter as an output instruction.

[0083] Specifically, the speed adjustment parameter of the turbine in the preset time period is determined according to the preset water flow fluctuation data, and the speed adjustment parameter is used as an output instruction to control the turbine for adjustment, including: obtaining the preset water flow speed of each preset time period in the preset water flow fluctuation data, and fitting the linear axis between the preset water flow speeds of each preset time period according to the least square method. The speed difference is mapped to the linear axis, the speed adjustment amount of each preset water flow speed is obtained, and the speed adjustment amount is converted into the adjustment parameter. The speed adjustment parameter is used as an output instruction to control the turbine for adjustment.

[0084] Specifically, when determining the adjustment parameter, it includes: obtaining the real-time response speed of the turbine speed, and determining whether to correct the adjustment parameter according to the relationship between the real-time response speed and the preset response speed: when the real-time response speed is equal to the preset response speed, it is determined that the adjustment parameter is not corrected. When the real-time response speed is less than the preset response speed, a correction coefficient is determined according to the response speed difference between the real-time response speed and the preset response speed, and the adjustment parameter is corrected according to the correction coefficient.

[0085] Specifically, when determining the correction coefficient according to the response speed difference between the real-time response speed and the preset response speed, it includes: determining the correction coefficient according to the relationship between the response speed difference and the preset first preset response speed difference and the second preset response speed difference. When the response speed difference is less than or equal to the first preset response speed difference, the correction coefficient is determined to be Q1. When the response speed difference is greater than the first preset response speed difference, and the response speed difference is less than or equal to the second preset response speed difference, the correction coefficient is determined to be Q2. When the response speed difference is greater than the second preset response speed difference, the correction coefficient is determined to be Q3. Among them, the first preset response speed difference is less than the second preset response speed difference, and 1<Q1<Q2<Q3<1.5.

[0086] It can be seen that the difference between the real-time speed and the preset speed is determined by comparing the real-time water flow data with the preset water flow fluctuation data. When the water flow fluctuation is consistent, unnecessary calculation and adjustment steps are avoided by directly taking the preset speed as the control target of the turbine, thereby improving the response speed. This method can reduce energy waste and over-adjustment. At this time, the preset water flow fluctuation data serves as a stable reference benchmark, and the optimal operation under minimum fluctuation is achieved by quickly adjusting the turbine on this benchmark. However, if there is an inconsistency between the real-time water flow data and the preset water flow fluctuation data, the dynamic adjustment is triggered by calculating the speed difference. In this process, the speed adjustment of the turbine no longer depends on a single preset value, but is corrected according to real-time conditions. Specifically, the water body flow velocity and water body pressure in the real-time water flow data are compared with the preset data to obtain the flow velocity difference and pressure difference, and then an adjustment coefficient is derived based on these differences to adjust the speed of the turbine. The real-time feedback of the two key parameters of flow velocity and pressure greatly enhances the self-correction of the turbine speed in a complex water flow environment. In addition, in the further refinement of the speed adjustment, the least squares method is used to fit the linear trend of the water flow fluctuation data. By mapping the speed difference to the linear axis, the speed adjustment can be effectively converted into the actual speed adjustment parameter. This method ensures that the smooth changes of the water flow fluctuation can be effectively tracked and avoids the negative impact of sudden water flow changes. The mapped adjustment parameters are fed back to the turbine through control instructions to achieve refined speed regulation, thereby ensuring the highest matching degree between the turbine and the water flow fluctuation.

[0087] Finally, the adjustment parameters are further corrected by calculating the response speed difference to adapt to the actual operating conditions. Through this correction coefficient, its control logic can be adjusted to cope with the errors in actual operation. In particular, when the response speed difference is less than or equal to the preset response speed difference, the adjustment parameters are considered to be sufficiently accurate, so no correction is performed. On the contrary, if the response speed difference is large, a stronger correction coefficient will be generated to ensure that the turbine wheel speed can be accurately adjusted to avoid the performance degradation of the turbine. This response speed difference correction mechanism is divided into multiple levels (Q1, Q2, Q3), each level corresponds to a different response correction strength, ensuring that appropriate correction strategies can be adopted according to different deviation ranges. This multi-level correction mechanism enables the turbine to operate at the most appropriate speed under various water flow fluctuation conditions, thereby effectively ensuring that the hydropower system maintains the stability of the power generation voltage under various water flow fluctuation conditions.

[0088] In the above embodiment, by acquiring the environmental information and water flow information of the target water area in real time and using the pre-established water flow fluctuation model, the turbine speed can be dynamically adjusted under different water flow fluctuation conditions, thereby ensuring that the power generation system can maintain high stability and efficiency under any water flow conditions. In addition, by comparing the real-time water flow data with the preset water flow fluctuation data in real time, the uncertainty of water flow fluctuation can be dealt with. If the real-time water flow data is consistent with the preset fluctuation data, it can be adjusted directly according to the preset speed, avoiding unnecessary calculations and energy waste. When there is a difference in water flow fluctuation, the turbine speed will be automatically adjusted according to the difference, avoiding the problem of unstable turbine operation caused by excessive water flow fluctuation. This not only enhances the ability to respond quickly to water flow fluctuations, but also improves the flexibility and robustness of the hydropower generation system through an intelligent adjustment process. On the other hand, based on the data accumulated over a long period of operation, the water flow fluctuation model is gradually optimized to make it more and more adaptable to different water areas. By mining the regularity in the data, the adjustment strategy is gradually adjusted during the process of water flow changes, avoiding blind and inefficient operations in traditional methods. This data-based adjustment method not only helps the hydropower system to effectively cope with complex water flow characteristics and achieve dynamic real-time adjustment, but also improves the accuracy of turbine operation and the overall efficiency of the power generation system. Moreover, through this intelligent dynamic adjustment method, the hydropower system can automatically adjust when facing different water flow fluctuations, avoiding problems such as over-adjustment or slow response, ensuring accurate control of turbine speed, and reducing equipment loss and energy waste.

[0089] In another preferred embodiment based on the above embodiment, Figure 2 As shown, this embodiment provides a turbine speed dynamic adjustment system based on machine learning, including: an acquisition module and a turbine control module.

[0090] Specifically, the acquisition module is configured to acquire environmental information and water flow information of the target water area, and substitute the environmental information and water flow information into a pre-established water flow fluctuation model to acquire preset water flow fluctuation data of the target water area within a preset period. The turbine control module is electrically connected to the acquisition module, and the turbine control module is configured to acquire real-time water flow data of the target water area within a preset period, and compare the real-time water flow data with the preset water flow fluctuation data, wherein: if the real-time water flow data is consistent with the preset water flow fluctuation data, the turbine control module determines the preset speed corresponding to the preset water flow fluctuation data as the preset turbine speed. If the real-time water flow data is inconsistent with the preset water flow fluctuation data, the turbine control module determines the adjustment coefficient according to the data difference between the real-time water flow data and the preset water flow fluctuation data, adjusts the preset speed according to the adjustment coefficient, and determines the adjusted preset speed as the preset turbine speed. The turbine control module is further configured to obtain the real-time rotation speed of the turbine and the speed difference between the real-time rotation speed and the preset turbine speed, wherein: if the real-time water flow data is consistent with the preset water flow fluctuation data, the turbine control module determines the speed adjustment parameter of the turbine within the preset time period according to the preset water flow fluctuation data, and controls the turbine to adjust according to the speed adjustment parameter as an output instruction. If the real-time water flow data is inconsistent with the preset water flow fluctuation data, the turbine control module determines the speed adjustment parameter of the turbine according to the speed difference and controls the turbine to adjust according to the speed adjustment parameter as an output instruction.

[0091] It can be understood that the method and system for dynamic adjustment of turbine speed based on machine learning in the above-mentioned embodiments of the present invention have the same beneficial effects and will not be described in detail.

[0092] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0093] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0094] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0095] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0096] Finally, 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 above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for dynamic adjustment of turbine speed based on machine learning, characterized in that: include: Obtaining environmental information and water flow information of a target water area, and substituting the environmental information and water flow information into a pre-established water flow fluctuation model to obtain preset water flow fluctuation data of the target water area within a preset time period; Acquire real-time water flow data of the target water area within the preset time period, and compare the real-time water flow data with the preset water flow fluctuation data, wherein: If the real-time water flow data is consistent with the preset water flow fluctuation data, determining a preset speed corresponding to the preset water flow fluctuation data as a preset turbine speed; If the real-time water flow data is inconsistent with the preset water flow fluctuation data, an adjustment coefficient is determined according to the data difference between the real-time water flow data and the preset water flow fluctuation data, the preset speed is adjusted according to the adjustment coefficient, and the adjusted preset speed is determined as the preset turbine speed; The real-time rotation speed of the turbine is obtained, and the speed difference between the real-time rotation speed and the preset turbine speed is obtained, wherein: If the real-time water flow data is consistent with the preset water flow fluctuation data, the speed adjustment parameter of the turbine within a preset time period is determined according to the preset water flow fluctuation data, and the turbine is controlled to adjust according to the speed adjustment parameter as an output instruction; If the real-time water flow data is inconsistent with the preset water flow fluctuation data, the speed adjustment parameter of the turbine is determined according to the speed difference and the speed adjustment parameter is used as an output instruction to control the turbine for adjustment.

2. The method for dynamic adjustment of turbine speed based on machine learning according to claim 1, characterized in that: When pre-establishing the water flow fluctuation model, include: Acquire the environmental information, water flow information and turbine speed of each historical water flow fluctuation of the target water area, and establish the water flow fluctuation relationship according to the environmental information, water flow information and turbine speed of each historical water flow fluctuation; Obtaining the linear relationship and distance measurement between each water flow fluctuation relational expression, and establishing the water flow fluctuation model according to the linear relationship and distance measurement between each water flow fluctuation relational expression; The environmental information and water flow information of the target water area are substituted into the water flow fluctuation model to obtain preset water flow fluctuation data of the target water area within a preset time period and a preset rotation speed of the turbine.

3. The method for dynamic adjustment of turbine speed based on machine learning according to claim 2, characterized in that: When the water flow fluctuation model is established according to the linear relationship and distance measurement between the water flow fluctuation relationship equations, it includes: Establishing a distance matrix according to the distance metric, and iteratively merging each of the water flow fluctuation relationship equations according to the distance matrix; Obtaining each of the water flow fluctuation relationship equations after iterative merging, and determining it as a water flow fluctuation impact type; Acquire environmental characteristics and water flow characteristics in the water flow fluctuation influence type, and classify each of the water flow fluctuation relationship equations based on the environmental characteristics and water flow characteristics; The linear relationship between each of the water flow fluctuation relationship equations in the water flow fluctuation influence type is obtained, and the water flow fluctuation model is established according to the linear relationship between each of the water flow fluctuation influence types.

4. The method for dynamic adjustment of turbine speed based on machine learning according to claim 3, characterized in that: According to the linear relationship between the water flow fluctuation influence types, the water flow fluctuation model is established, including: Obtaining the environmental information and water flow information during any of the historical water flow fluctuations, and substituting the environmental information and water flow information during the historical water flow fluctuations into the water flow fluctuation model to obtain the preset water flow fluctuation data; Comparing the water flow fluctuation data of the historical water flow fluctuation pattern with the preset water flow fluctuation data, wherein: If the water flow fluctuation data is inconsistent with the preset water flow fluctuation data, the water flow fluctuation model is corrected.

5. The method for dynamic adjustment of turbine speed based on machine learning according to claim 1, characterized in that: If the real-time water flow data is inconsistent with the preset water flow fluctuation data, determining the adjustment coefficient according to the data difference between the real-time water flow data and the preset water flow fluctuation data includes: Acquire the water flow velocity and water pressure in the real-time water flow data, and preset the water flow velocity and preset water pressure in the preset water flow fluctuation data; Obtaining a flow velocity difference between the water body flow velocity and a preset water body flow velocity, and a pressure difference between the water body pressure and a preset water body pressure; According to the relationship between the flow rate difference and the pre-configured preset flow rate difference range, and the pressure difference and the pre-configured preset pressure difference range, an acquisition index of the adjustment coefficient is determined, and the adjustment coefficient is determined according to the acquisition index, wherein: If the flow rate difference is within the preset flow rate difference range, and the pressure difference is within the preset pressure difference range, then the acquisition index of the adjustment coefficient is determined to be zero, and the adjustment coefficient is 0; If the flow velocity difference is within the preset flow velocity difference range, and the pressure difference is not within the pressure difference range, determining that the acquisition index of the adjustment coefficient is the pressure difference; If the flow velocity difference is not within the preset flow velocity difference range, and the pressure difference is within the preset pressure difference range, then determining that the acquisition index of the adjustment coefficient is the flow velocity difference; If the flow rate difference is not within the preset flow rate difference range, and the pressure difference is not within the preset pressure difference range, then the acquisition index of the adjustment coefficient is determined to be a combination of the flow rate difference and the pressure difference.

6. The method for dynamic adjustment of turbine speed based on machine learning according to claim 5, characterized in that: When determining the adjustment coefficient according to the acquisition index, it includes: Substitute the acquisition index into the formula to obtain the adjustment coefficient, where the formula is as follows: ； Among them, S is the adjustment coefficient, 1 is the weight coefficient, M is the acquisition index, and M min To obtain the minimum value of the indicator, M max To obtain the maximum value of the indicator.

7. The method for dynamic adjustment of turbine speed based on machine learning according to claim 1, characterized in that: Determining the speed adjustment parameter of the turbine within a preset time period according to the preset water flow fluctuation data, and controlling the turbine to adjust according to the speed adjustment parameter as an output instruction, including: Obtaining a preset water flow velocity in each preset time period in the preset water flow fluctuation data, and fitting a linear axis between the preset water flow velocities in each preset time period according to a least squares method; Mapping the speed difference to the linear axis, obtaining the speed adjustment amount of each of the preset water flow speeds, and converting the speed adjustment amount into the adjustment parameter; The turbine is controlled to be adjusted according to the speed adjustment parameter as an output instruction.

8. The method for dynamic adjustment of turbine speed based on machine learning according to claim 7, characterized in that: When determining the adjustment parameters, it includes: Acquiring a real-time response speed of the turbine speed, and determining whether to modify the adjustment parameter according to a relationship between the real-time response speed and a preset response speed; When the real-time response speed is equal to the preset response speed, determining not to modify the adjustment parameter; When the real-time response speed is less than the preset response speed, a correction coefficient is determined according to a response speed difference between the real-time response speed and the preset response speed, and the adjustment parameter is corrected according to the correction coefficient.

9. The method for dynamic adjustment of turbine speed based on machine learning according to claim 8, characterized in that: Determining the correction coefficient according to the response speed difference between the real-time response speed and the preset response speed includes: Determining the correction coefficient according to a relationship between the response speed difference and a preset first response speed difference and a preset second response speed difference; When the response speed difference is less than or equal to the first preset response speed difference, the correction coefficient is determined to be Q1; When the response speed difference is greater than the first preset response speed difference, and the response speed difference is less than or equal to the second preset response speed difference, the correction coefficient is determined to be Q2; When the response speed difference is greater than the second preset response speed difference, the correction coefficient is determined to be Q3; The first preset response speed difference is smaller than the second preset response speed difference, and 1<Q1<Q2<Q3<1.

5.

10. A turbine speed dynamic adjustment system based on machine learning, using a turbine speed dynamic adjustment method based on machine learning as claimed in any one of claims 1 to 9, characterized in that: include: An acquisition module is configured to acquire environmental information and water flow information of a target water area, and substitute the environmental information and water flow information into a pre-established water flow fluctuation model to acquire preset water flow fluctuation data of the target water area within a preset time period; a turbine control module, electrically connected to the acquisition module, the turbine control module being configured to acquire real-time water flow data of the target water area within the preset time period, and compare the real-time water flow data with the preset water flow fluctuation data, wherein: if the real-time water flow data is consistent with the preset water flow fluctuation data, the turbine control module determines the preset speed corresponding to the preset water flow fluctuation data as the preset turbine speed; if the real-time water flow data is inconsistent with the preset water flow fluctuation data, the turbine control module determines an adjustment coefficient according to a data difference between the real-time water flow data and the preset water flow fluctuation data, adjusts the preset speed according to the adjustment coefficient, and determines the adjusted preset speed as the preset turbine speed; Wherein, the turbine control module is also configured to obtain the real-time rotation speed of the turbine, and obtain the speed difference between the real-time rotation speed and the preset turbine speed, wherein: if the real-time water flow data is consistent with the preset water flow fluctuation data, the turbine control module determines the speed adjustment parameter of the turbine within a preset time period according to the preset water flow fluctuation data, and controls the turbine to adjust according to the speed adjustment parameter as an output instruction; if the real-time water flow data is inconsistent with the preset water flow fluctuation data, the turbine control module determines the speed adjustment parameter of the turbine according to the speed difference, and controls the turbine to adjust according to the speed adjustment parameter as an output instruction.

Images