Partition modeling method, device, program product and electronic equipment for water pump turbine

Abstract

This disclosure relates to a method, apparatus, program product, and equipment for partitioned modeling of a water pump turbine. The method includes: acquiring a first state variable and a second state variable of the water pump turbine; determining a predicted value of the second state variable based on the first state variable; determining the error between the actual value and the predicted value; determining a first value for the error parameter; partitioning the target state variable in the first state variable according to the first value, initially determining multiple local intervals; if the error of any local interval exceeds an error threshold, introducing an intermediate variable and determining a second value for the intermediate variable; determining a third value; dividing the water pump turbine into multiple partitions based on the first, second, and third values; determining a comprehensive region model corresponding to the water pump turbine in each partition; and determining the parameter values ​​of the first state variable and the target value of the second state variable of the water pump turbine based on the comprehensive region model to control the water pump turbine. This disclosure enables precise control of the water pump turbine.

Description

Technical Field

[0001] This disclosure relates to the field of water pump turbine technology, and more specifically, to a method for partitioning and modeling water pump turbines, a device for partitioning and modeling water pump turbines, a computer program product, and electronic equipment. Background Technology

[0002] As a core component of pumped storage systems, the pump-turbine exhibits significant nonlinearity, strong coupling, and an "S" characteristic region in its dynamic characteristics, meaning that multiple unit flow rates or unit torque values ​​exist per unit speed. Existing turbine modeling methods in related technologies struggle to accurately describe the system's dynamic behavior, resulting in poor control accuracy, low model interpretability, and low computational efficiency. Summary of the Invention

[0003] The purpose of this disclosure is to provide a method for partitioned modeling of water pump turbines, a device for partitioned modeling of water pump turbines, a computer program product, and electronic equipment, thereby overcoming, to at least a certain extent, the problems of poor control accuracy and low computational efficiency caused by the limitations and defects of related technologies.

[0004] According to one aspect of this disclosure, a method for partitioning a pump-turbine is provided, comprising:

[0005] Obtain the first and second state variables from the experimental data of the water pump turbine, and determine the predicted value of the second state variable based on the first state variable;

[0006] Determine the error between the actual value and the predicted value of the second state variable, determine the first value of the parameter corresponding to the error, and partition the target state variable in the first state variable according to the first value to initially determine multiple local intervals.

[0007] If the error in any local interval is greater than the error threshold, an intermediate variable is introduced and its second value is determined.

[0008] Determine the third value corresponding to the specified state variable in the second state variable, divide multiple partitions according to the first value, the second value and the third value, and determine the comprehensive area model corresponding to the water pump and water turbine in each partition;

[0009] Based on the comprehensive regional model, the parameter values ​​of the first state variable and the target value of the second state variable of the water pump turbine are determined, so as to control the water pump turbine according to the parameter values ​​and the target values.

[0010] In one exemplary embodiment of this disclosure, determining the predicted value of the second state variable based on the first state variable includes:

[0011] The first state variable and the second state variable are combined into a data matrix, and the first matrix and the second matrix are determined based on the data matrix;

[0012] By combining the observation function set, the first matrix and the second matrix are established as a linear relationship to generate a linear relationship between the first state variable and the second state variable;

[0013] Based on the initial values ​​and linear relationship of the first and second state variables, the predicted values ​​of the second state variables in other states are determined.

[0014] In one exemplary embodiment of this disclosure, the step of combining the observation function set to establish a linear relationship between the first matrix and the second matrix to generate a linear relationship between the first state variable and the second state variable includes:

[0015] Using the set of observation functions, a linear relationship is established between the first matrix and the second matrix, which is then transformed into an objective equation, which is a constant matrix.

[0016] The first matrix is ​​decomposed to obtain the component form. Based on the component form, the constant matrix is ​​transformed. The transformed constant matrix is ​​then decomposed by eigenvalues ​​to obtain the decomposed form.

[0017] The transformed constant matrix is ​​combined with the decomposition form to generate a linear relationship between the first state variable and the second state variable.

[0018] In one exemplary embodiment of this disclosure, the step of introducing an intermediate variable within the failure interval to determine a second value of the intermediate variable includes:

[0019] An intermediate variable is introduced within the failure interval, and a data matrix is ​​reconstructed based on the intermediate variable. The linear relationship between the intermediate variable and the second state variable is then determined again based on the reconstructed data matrix.

[0020] Based on the re-established linear relationship, the predicted value of the second state variable is determined, and the second value of the intermediate variable is determined based on the error between the predicted value and the actual value.

[0021] In one exemplary embodiment of this disclosure, determining the integrated region model corresponding to the pump turbine in each partition includes:

[0022] Within each partition, determine the partition fitting model for each second state variable composed of multiple first state variables;

[0023] By fitting different second-state variables to partitioned models and combining them according to the same partition, the comprehensive regional model corresponding to the water pump turbine in each partition is determined.

[0024] In one exemplary embodiment of this disclosure, the determination of the first value of the parameter corresponding to the error includes:

[0025] Gradually increase the value of the parameter until the calculated error is greater than the error threshold, and then determine the first value of the parameter.

[0026] In an exemplary embodiment of this disclosure, the first state variable includes the guide vane opening and unit rotational speed of the water pump turbine, the second state variable includes the flow rate and torque of the water pump turbine, and the intermediate variable is related to the target state variable and the second state variable.

[0027] According to one aspect of this disclosure, a partitioning modeling apparatus for a water pump turbine is provided, comprising:

[0028] The prediction value determination module is used to acquire the first state variable and the second state variable from the experimental data of the water pump turbine, and determine the predicted value of the second state variable based on the first state variable.

[0029] The initial partitioning module is used to determine the error between the actual value and the predicted value of the second state variable, determine the first value of the parameter corresponding to the error, and partition the target state variable in the first state variable according to the first value to initially determine multiple local intervals.

[0030] The intermediate variable introduction module is used to introduce an intermediate variable and determine its second value if the error in any local interval is greater than the error threshold.

[0031] The model building module is used to determine the third value corresponding to the specified state variable in the second state variable, divide multiple partitions according to the first value, the second value and the third value, and determine the comprehensive area model corresponding to the water pump and water turbine in each partition.

[0032] The control module is used to determine the parameter values ​​of the first state variable and the target value of the second state variable of the pump turbine based on the integrated regional model, so as to control the pump turbine according to the parameter values ​​and the target value.

[0033] According to one aspect of this disclosure, a computer program product is provided, which, when executed by a processor, implements the partition modeling method for a water pump turbine as described above.

[0034] According to one aspect of this disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to implement the pump-turbine partitioning modeling method described above by executing the executable instructions.

[0035] In the technical solution provided in this disclosure, on the one hand, a comprehensive region model of the pump-turbine is constructed based on the first and second state variables. This comprehensive region model allows for control of the pump-turbine's first and second state variables. Since the target value of the second state variable can be accurately calculated based on the first state variable, or the first state variable can be adjusted based on the second state variable, the control accuracy of the pump-turbine can be improved. On the other hand, since the target value of the second state variable can be uniquely determined based on the comprehensive region model, and the process of further judgment from multiple values ​​is avoided, processing efficiency and model interpretability are improved.

[0036] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0037] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0038] Figure 1 The schematic diagram illustrates a flow chart of a partition modeling method for a water pump turbine according to an embodiment of the present disclosure.

[0039] Figure 2 The schematic diagram illustrates the flow chart of the artificial intelligence algorithm in an embodiment of this disclosure.

[0040] Figure 3 The diagram illustrates the error between flow rate and guide vane opening in an embodiment of this disclosure.

[0041] Figure 4 The diagram illustrates the error between flow rate and unit rotation speed in an embodiment of this disclosure.

[0042] Figure 5 The diagram illustrates the error between flow rate and intermediate variables in an embodiment of this disclosure.

[0043] Figure 6 A schematic diagram illustrating the partitioning of traffic in an embodiment of this disclosure is shown.

[0044] Figure 7 The diagram illustrates the error between torque and guide vane opening in an embodiment of this disclosure.

[0045] Figure 8 The diagram illustrates the error between torque and unit speed in an embodiment of this disclosure.

[0046] Figure 9 The diagram illustrates the error between torque and intermediate variables in an embodiment of this disclosure.

[0047] Figure 10 A schematic diagram illustrating the torque partitioning in an embodiment of this disclosure is shown.

[0048] Figure 11 The schematic diagram illustrates the integrated zoning of the water pump turbine in an embodiment of this disclosure.

[0049] Figure 12 The schematic diagram illustrates a block diagram of a partition modeling device for a water pump turbine in an embodiment of this disclosure.

[0050] Figure 13 A schematic block diagram of an electronic device according to an embodiment of the present disclosure is shown. Detailed Implementation

[0051] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this disclosure more comprehensive and complete, and to fully convey the concept of the example embodiments to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more of the specific details omitted, or other methods, components, apparatus, steps, etc., can be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0052] Furthermore, the accompanying drawings are merely illustrative of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0053] As the core component of pumped-storage systems, pump-turbines exhibit significant nonlinearity, strong coupling, and an "S"-shaped characteristic region in their dynamic characteristics, meaning they possess multiple unit flow rates or unit torque values ​​per unit rotational speed. Therefore, traditional turbine modeling methods struggle to accurately describe the system's dynamic behavior, particularly in terms of control accuracy, model interpretability, and computational efficiency. Methods such as analytical formulas, characteristic curve interpolation, or black-box neural network models cannot simultaneously achieve both accuracy and interpretability.

[0054] To address the aforementioned technical problems, this disclosure provides a method for zonal modeling of pump-turbines. This method combines artificial intelligence algorithms with experimental data to automatically construct a multi-interval explicit model. This method can be applied to zonal modeling of pump-turbines in pumped storage systems. Based on the established comprehensive regional model, the flow rate or torque of the pump-turbine can be accurately determined according to the guide vane opening and unit rotational speed, thus improving the accuracy of the determined flow rate or torque. Figure 1 The diagram illustrates a partitioned modeling method for a water pump turbine. (See reference...) Figure 1 As shown, the method mainly includes the following steps:

[0055] Step S110: Obtain the first state variable and the second state variable from the experimental data of the water pump turbine, and determine the predicted value of the second state variable based on the first state variable;

[0056] Step S120: Determine the error between the actual value and the predicted value of the second state variable, determine the first value of the parameter corresponding to the error, and partition the target state variable in the first state variable according to the first value to initially determine multiple local intervals.

[0057] Step S130: If the error in any local interval is greater than the error threshold, introduce an intermediate variable and determine the second value of the intermediate variable.

[0058] Step S140: Divide the first state variable into multiple partitions based on the first value and the second value, and determine the comprehensive region model corresponding to the water pump and water turbine in each partition.

[0059] Step S150: Based on the integrated regional model, determine the parameter values ​​of the first state variable and the target values ​​of the second state variable of the water pump turbine, so as to control the water pump turbine according to the parameter values ​​and the target values.

[0060] Next, the method for partitioning the water pump turbine in the present disclosure will be described in detail with reference to the embodiments.

[0061] Step S110: Obtain the first state variable and the second state variable from the experimental data of the water pump turbine, and determine the predicted value of the second state variable based on the first state variable.

[0062] In this embodiment, the pump-turbine can be a core component of a pumped-storage system, and can be of any structure and with any head. Experimental data refers to data generated by the experimental platform, which may include multiple state variables, specifically a first state variable and a second state variable. These can include the guide vane opening α and the unit rotational speed n of the pump-turbine. 11 The flow rate Q of the water pump turbine is used as the first state variable. 11 And the torque M of the water pump and turbine 11 As the second state variable, flow rate can be per unit flow rate, and torque refers to per unit torque. The first state variable can be used to represent the independent variable, and the second state variable can be used to represent the dependent variable.

[0063] In some embodiments, the first state variable can be represented as S (u) The second state variable can be represented as S. (v) In the experimental data, all data for the first state variable with respect to the second state variable were collected and organized into a data matrix. This data matrix can be represented as DM.

[0064] In some embodiments, a set of observation functions can be selected based on relevant experience. The set of observation functions can be the first, square, or cube power of the variable. The selected set of observation functions can be applied to the data matrix to perform a dimensionality increase operation, resulting in a high-dimensional data matrix K. Next, a first matrix and a second matrix can be selected from the high-dimensional data matrix K. The first and second matrices can have the same length, for example, both having N-1 columns, but their specific values ​​can be different. For example, the first matrix K... a A matrix can be generated from the data in the first, second, and third columns of a high-dimensional data matrix K, and the second matrix K is... b It can generate a matrix from the data in the 2nd, 3rd, and 4th columns of a data matrix K in a high-dimensional space.

[0065] After determining the first and second matrices, it can be assumed that the observation function set used above can establish a linear relationship between the first and second matrices. Furthermore, the established linear relationship can be transformed to obtain the objective equation as shown in formula (1):

[0066]

[0067] Where A is a constant matrix. It is K a The false rebellion.

[0068] Furthermore, the first matrix can be decomposed to obtain its component form. For example, the first matrix K can be decomposed into its component form. a Performing SVD decomposition yields its component form. Correspondingly, the constant matrix A represented by the objective equation can be transformed into the following form:

[0069] A = K b ·V(Σ) -1 (U) * Formula (2)

[0070] Next, the constant matrix A is decomposed into eigenvalues, resulting in the following decomposition form:

[0071]

[0072] Where W is the eigenvector matrix of the constant matrix A, and Λ is the eigenvalue matrix.

[0073] Substituting formula (2) into formula (3), we get formula (4). Formula (4) is used to represent the linear relationship between state variables, that is, the linear relationship between the first state variable and the second state variable.

[0074]

[0075] Using the initial values ​​of the first and second state variables, and the aforementioned linear relationship, the predicted values ​​for other states can be calculated. The predicted values ​​for other states refer to the predicted values ​​of the second state variable under other conditions. Other states refer to states different from their initial values. Based on this linear relationship, the value of any state from the first state variable to the second state variable can be linearly predicted.

[0076] In step S120, the error between the actual value and the predicted value of the second state variable is determined, the first value of the parameter corresponding to the error is determined, and the target state variable in the first state variable is partitioned according to the first value to initially determine multiple local intervals.

[0077] In this embodiment of the disclosure, after determining the predicted value of the second state variable, the error between the predicted value and the actual value can be calculated. For example, the error can be determined based on the difference between the processing results corresponding to the predicted value and the actual value. The processing result corresponding to the actual value can be determined based on a linear relationship, an eigenvalue matrix, and the product of the actual values. It should be noted that the error may include a parameter N. When determining the parameter N, the value of parameter N can be gradually increased, and the error can be recalculated based on the value of N until the calculated error is greater than an error threshold. The value of parameter N corresponding to the error being greater than the error threshold is taken as the first value of parameter N. The error between the predicted value and the actual value can be expressed as:

[0078]

[0079] After obtaining the error between the predicted and actual values, the error can be used to trigger partitioned modeling. For example, the target state variable in the first state variable can be partitioned according to a first value of parameter N. The target state variable can be a unit rotational speed n. 11 Based on the unit rotational speed, multiple local intervals can be initially determined, and the dividing points of these local intervals can be determined according to the first value of parameter N. For example, when the unit rotational speed is greater than 0, it can be seen from the error curve of the unit rotational speed that the first value of parameter N is 73. Therefore, multiple local intervals can be divided according to the first value of parameter N, 73.

[0080] In step S130, if the error in any local interval is greater than the error threshold, an intermediate variable is introduced and a second value of the intermediate variable is determined.

[0081] For example, the error curve can first be scanned and analyzed. If the error in a certain local interval of the error curve exceeds a preset threshold, the fitting of that local interval is determined to be invalid. Based on this, intermediate variables can be introduced into the invalid region, and the dimensionality mapping and local modeling can be re-performed based on the intermediate variables. The polynomial model is solved independently for each sub-interval, and the accuracy is verified by the error function to ensure the continuity and accuracy of the modeling.

[0082] For example, an intermediate variable can be introduced within the failure interval, the data matrix can be reconstructed based on the intermediate variable, and the linear relationship between the intermediate variable and the second state variable can be determined again based on the reconstructed data matrix; the predicted value of the second state variable can be determined according to the reconstructed linear relationship, and the second value of the intermediate variable can be determined when the error between the predicted value and the actual value is a critical value.

[0083] In this embodiment of the disclosure, what is to be constructed is Q 11 with a,n 11 The relationship between them, and M 11 with a,n 11The relationship between them. In the process of constructing different relationships, the intermediate variables introduced in the failure interval can be related to the target state variable and the second state variable to which the relationship is to be established. The target state variable can be a unit rotational speed, and the second state variable to which the relationship is to be established can be a unit flow rate or a unit torque.

[0084] For example, for the target state variable in the first state variable, when the second state variable is flow rate, the target state variable in the first state variable can be partitioned according to the first value of the parameter corresponding to the error, initially determining multiple local intervals corresponding to the target state variable. The parameter corresponding to the error is the same as the target state variable. When the target state variable is unit rotational speed, the initially determined multiple local intervals can be divided according to the first value of the parameter corresponding to the error. For example, when the first value of the parameter is 73, the first partition can be determined based on the first value of the parameter. For example, 0≤n 11 ≤73 is defined as Q 11 —n 11 The first partition.

[0085] In constructing Q 11 with a,n 11 When there is a relationship between n 11 Within a local interval ≥73, where the error exceeds the error threshold, this local interval can be defined as the failure interval to continue searching for a linear relationship. An intermediate variable is then introduced within this failure interval. This intermediate variable can be related to the target state variable and the second state variable with which the relationship is to be established; for example, it can be related to unit rotational speed and flow rate. The error is further calculated based on the introduced intermediate variable, and a second value for the intermediate variable is determined when the error curve approaches the boundary value. Specifically, the intermediate variable can be represented as:

[0086]

[0087] In constructing M 11 with a,n 11 When considering the relationship between the target state variable in the first state variable and the torque of the water pump turbine in the second state variable, the target state variable in the first state variable can be partitioned according to the first value of the parameter corresponding to the error, initially determining multiple local intervals corresponding to the target state variable. The parameter corresponding to the error is the same as the target state variable. When the target state variable is unit speed, the initially determined multiple local intervals can be divided according to the first value of the parameter corresponding to the error. For example, when the first value of the parameter is 73, the first partition can be determined based on the first value of the parameter. For example, 0≤n 11 ≤73 is defined as Q 11 —n 11 The first partition.

[0088] If n 11 Within a local interval ≥73, the error exceeds the error threshold. To continue searching for a linear relationship, this local interval can be defined as the failure interval, and an intermediate variable can be introduced within it. This intermediate variable can be related to the target state variable and the second state variable with which the relationship is to be established; for example, it can be related to unit speed and the torque of the water pump turbine. The error is then calculated based on the introduced intermediate variable, and a second value for the intermediate variable is determined when the error curve approaches the boundary value. Specifically, the intermediate variable can be represented as:

[0089]

[0090] In step S140, a third value corresponding to a specified state variable in the second state variable is determined. Multiple partitions are divided according to the first value, the second value, and the third value, and a comprehensive regional model corresponding to the water pump and turbine in each partition is determined.

[0091] In this embodiment of the disclosure, a third value of a specified state variable in the second state variable can be determined. The specified state variable can be the unit torque of the water pump turbine, and the third value can be 0. It should be noted that when determining Q... 11 with a,n 11 The relationship between them and the determination of M 11 with a,n 11 When defining the relationship between them, the specified state variable can be the unit torque of the water pump turbine.

[0092] After determining the first, second, and third values, the first partition can be determined based on the first value of the parameter, the second partition can be determined based on the first value of the parameter and the second value of the intermediate variable, and the third partition can be determined based on the second value of the intermediate variable and the third value of the specified state variable.

[0093] For example, in determining Q 11 with a,n 11 When discussing the relationship between parameters, if the first value of the parameter is 73, then 0 ≤ n can be used. 11 ≤73 is defined as Q 11 —n 11 The first partition. When the intermediate variable IVQ is 79, n can be... 11 The region where Q > 73 and IVQ ≤ 79 is defined as Q. 11 —n 11 The second partition. When IVQ > 79, use M. 11 =0 is used as the dividing line, because in M 11 A value less than 0 will result in a severe shift in control. Therefore, IVQ > 79 and M... 11 The region > 0 is defined as Q 11 —n 11The third partition.

[0094] Similarly, in determining M 11 with a,n 11 When discussing the relationship between parameters, if the first value of the parameter is 73, then 0 ≤ n can be used. 11 ≤73 is defined as M 11 —n 11 The first partition. For example, when the intermediate variable IVM takes the value of 67, n can be... 11 The region with IVM > 73 and IVM ≤ 67 is defined as M. 11 —n 11 The second partition. When IVM > 67, use M. 11 =0 is used as the dividing line, because in M 11 A value less than 0 will lead to a severe shift in control. Therefore, IVM > 67 and M 11 The region > 0 is defined as M 11 —n 11 The third partition.

[0095] After identifying multiple zones, a comprehensive regional model corresponding to the pump-turbine in each zone can be determined. For example, firstly, a zone-fitting model for each second state variable with respect to multiple first state variables is determined within each zone. The zone-fitting model can include the influence of each first state variable on the second state variables; that is, the dependent variable in the zone-fitting model can be a second state variable, and the independent variable can be each first state variable. For example, a zone-fitting model can be constructed for each first state variable regarding flow rate in different zones, and a zone-fitting model can be constructed for each first state variable regarding the torque of the pump-turbine in different zones.

[0096] Create Q in each partition 11 and M 11 The polynomial model is used as the partition fitting model, where the terms of the polynomial are functions contained in the set of observation functions.

[0097] For example, the partition fitting model for the traffic established in the first partition can be as shown in formula (8):

[0098]

[0099] The partition fitting model for the flow established in the second partition can be shown in formula (9):

[0100]

[0101] The partition fitting model for the flow established in the third partition can be shown in Equation (10):

[0102]

[0103] The partitioned fitting model for the torque of the pump turbine established in the first partition can be shown in Equation (11):

[0104]

[0105] The partitioned fitting model for the torque of the pump turbine established in the second partition can be found in formula (12):

[0106]

[0107] The partitioned fitting model for the torque of the pump-turbine established in the third partition can be found in formula (13).

[0108]

[0109] Q in the partition fitting model 11 The coefficients are shown in Table 1, and the partition fitting model M 11 The coefficients are shown in Table 2.

[0110] Table 1

[0111]

[0112] Table 2

[0113]

[0114] Furthermore, partitioned fitting models with different second-state variables can be combined according to the same partition to determine the comprehensive regional model corresponding to the pump-turbine in each partition. Since Q 11 and M 11 The partitions are interleaved and overlapping. Furthermore, by combining the partitions of the two types of partitions, a comprehensive regional model of the water turbine is obtained.

[0115] Figure 2 The flowchart of an artificial intelligence algorithm is illustrated in the diagram. Figure 2 As shown, the main steps include:

[0116] Step S202: Obtain experimental data.

[0117] Step S204: Preprocess the experimental data.

[0118] Step S206: Perform the dimensional upgrade operation.

[0119] Step S208: Establish the objective equation.

[0120] Step S210: Calculate the predicted value of the second state variable.

[0121] Step S212: Calculate the error between the predicted value and the actual value.

[0122] Step S214: Determine whether the error is less than the error threshold; if yes, proceed to step S216; if no, proceed to step S218.

[0123] Step S216, N = N + 1, and proceed to step S210.

[0124] Step S218: Determine parameter N.

[0125] Step S220: Obtain the fitted model.

[0126] In this embodiment, a pumped storage power station collected experimental data on its pump-turbine under medium head (413m) conditions, with guide vane opening ranging from 0.4 to 1.0 and unit rotational speed from 0 to 1.8. The following steps can then be used for modeling:

[0127] A data matrix X = [a, n] can be constructed. 11 ], Target variable Y = [Q 11 M 11 ], where the target variable can be used to represent the second state variable. The aforementioned artificial intelligence algorithm can be used to establish Q. 11 -a, Q 11 -n 11 M 11 -a and M 11 -n 11 The relationship is then used as a basis to build a model.

[0128] First, regarding the construction of Q 11 with a,n 11 The specific process of the relationship between them will be explained.

[0129] Based on the above method of calculating errors, the highest power observation function set can be selected to process the data according to the empirical relationship between the flow rate of the water pump turbine and the guide vane opening. The error curve obtained by using artificial intelligence algorithms is shown below. Figure 3 As shown. Based on the empirical relationship between the flow rate and unit speed of a water pump turbine, a set of observation functions with a first power was selected. As the unit speed changes, the linear relationship of the high-dimensional control also changes; the specific steps for partitioning it include:

[0130] When the unit speed is greater than zero, the error curve is as follows: Figure 4 As shown, by Figure 4 As can be seen, the parameter refers to the unit rotational speed, and its value is the first value, 73. Based on this, 0≤n can be used. 11 ≤73 is defined as Q 11 —n 11 The first partition.

[0131] When n 11 When the value is ≥73, to continue searching for a linear relationship, an intermediate variable IVQ can be introduced within the failure region. This intermediate variable is related to unit rotational speed and flow rate. When the intermediate variable is introduced, the error curve obtained using an artificial intelligence algorithm is as follows: Figure 5 As shown. From Figure 5 It can be seen that when the error approaches the critical point, the parameter N (i.e., IVQ) takes the value of 79, therefore the second value of the intermediate variable is 79. Based on this, n 11 The region where Q > 73 and IVQ ≤ 79 is defined as Q. 11 —n 11 The second partition.

[0132] When IVQ > 79, use M 11 =0 is used as the dividing line, because in M 11 A value less than 0 will result in a severe control shift. Therefore, setting IVQ > 79 and M... 11 The region > 0 is defined as Q 11 —n 11 The third partition.

[0133] Based on the above three scenarios, Q will be available within the specified range. 11 with a,n 11 The relationships between them are divided into three partitions, and these three partitions can be represented as follows: Figure 6 As shown.

[0134] Next, we will construct M. 11 with a,n 11 The specific process of the relationship between them will be explained.

[0135] Based on the empirical relationship between turbine torque and guide vane opening, the highest second power observation function set is selected to process the data. The error curve obtained using artificial intelligence algorithms is shown below. Figure 7 As shown. Based on the empirical relationship between turbine torque and unit speed, a set of cubic power observation functions is selected.

[0136] When the unit speed is greater than zero, the error curve is as follows: Figure 8 As shown. From Figure 8 As can be seen from this, the parameter N (i.e., the unit rotational speed n) 11 The value is 73, meaning the first value is 73. Based on this, 0 ≤ n 11 ≤73 is defined as M 11 —n 11 The first partition.

[0137] When n 11When the value is ≥73, an intermediate variable IVM can be introduced within the failure range. This intermediate variable is related to the unit speed and the torque of the pump-turbine. After introducing the intermediate variable, the error curve obtained using the artificial intelligence algorithm is as follows: Figure 9 As shown, the parameter N (i.e., the intermediate variable IVM) has a value of 67 at this point, meaning the second value of the intermediate variable is 67. Based on this, n... 11 The region with IVM > 73 and IVM ≤ 67 is defined as M. 11 —n 11 The second partition.

[0138] When IVM > 67, the third value of the specified state variable, i.e., the torque of the pump-turbine, can be determined. The third partition is then determined based on the second value of the intermediate variable and the third value of the specified state variable. For example, if IVM > 67 and M... 11 The region > 0 is defined as M 11 —n 11 The third partition.

[0139] Based on the above three scenarios, M will be used within the available range. 11 with a,n 11 The relationships between them are divided into three partitions, such as Figure 10 As shown.

[0140] Traffic Q 11 The torque M of the water pump and turbine 11 The partitions are overlapping. Furthermore, different partition fitting models for the same partition can be combined; for example, multiple partition fitting models for the first partition can be combined, multiple partition fitting models for the second partition can be combined, and so on, thus obtaining a comprehensive region model of the pump-turbine. Figure 11 As shown. Based on Figure 11 As shown, the partitioning conditions and the selection of model formulas are as follows:

[0141] The condition for region 1 is: M 11 >0, n 11 ≤73, IVM≤67, the integrated region model is: Q 11 =QP1,M 11 =MP1.

[0142] The condition for region 2 is: M 11 >0, n 11 >73, IVM≤67, a≤a*. The comprehensive region model is: Q 11 =QP2,M 11 =MP2.

[0143] The condition for region 3 is: M 11 >0, n 11>73, IVM>67, a≤a*. The comprehensive region model is: Q 11 =QP2,M 11 =MP3.

[0144] The condition for region 4 is: M 11 >0, n 11 >73, IVQ≤79, a>a*. The comprehensive region model is: Q 11 =QP2,M 11 =MP2.

[0145] The condition for region 5 is: M 11 >0, n 11 >73, IVQ≤79, IVM≤67. The comprehensive region model is: Q 11 =QP3,M 11 =MP2.

[0146] The condition for region 6 is: M 11 >0, n 11 >73, IVQ>79, IVM>67. The comprehensive region model is: Q 11 =QP3,M 11 =MP3.

[0147] In step S150, the parameter values ​​of the first state variable and the target values ​​of the second state variable of the water pump turbine are determined based on the integrated regional model, so as to control the water pump turbine according to the parameter values ​​and the target values.

[0148] In this embodiment, after constructing a comprehensive region model of the pump-turbine, a value for a first state variable can be given. Based on the given first state variable and the constructed comprehensive region model, the target value of a second state variable corresponding to the current value can be accurately searched. This avoids the situation where multiple values ​​are found based on a single first state variable, thus improving the accuracy and uniqueness of the data search. Furthermore, since a comprehensive region model exists between the second and first state variables, the parameter values ​​of the first state variables can be precisely adjusted based on the second state variables. This allows for precise control of the pump-turbine's operation based on the parameter values ​​of the first and the target value of the second state variables. Therefore, the system dynamic behavior of the pump-turbine can be accurately described, improving the control precision of the pump-turbine, increasing model interpretability, and enhancing computational efficiency.

[0149] In this embodiment, a multi-interval explicit model is automatically constructed using experimental data based on an upgraded observation function set and a linear mapping mechanism. The modeling using an artificial intelligence-based upgraded mechanism is highly accurate and can describe the nonlinear multi-valued relationships in the "S" characteristic region. The model employs a polynomial explicit expression, facilitating understanding, optimization, and integration into the controller. It is applicable to pumps and turbines with different structures and water heads, and avoids the overfitting and uninterpretable problems associated with black-box neural network models.

[0150] In this embodiment of the disclosure, a partitioning modeling device for a water pump turbine is also provided, with reference to... Figure 12 As shown, the partition modeling device 1200 for the water pump turbine includes:

[0151] The prediction value determination module 1201 is used to acquire the first state variable and the second state variable in the experimental data of the water pump turbine, and determine the predicted value of the second state variable based on the first state variable.

[0152] The initial partitioning module 1202 is used to determine the error between the actual value and the predicted value of the second state variable, determine the first value of the parameter corresponding to the error, and partition the target state variable in the first state variable according to the first value to initially determine multiple local intervals.

[0153] The intermediate variable introduction module 1203 is used to introduce an intermediate variable and determine the second value of the intermediate variable if the error in any local interval is greater than the error threshold.

[0154] The model building module 1204 is used to determine the third value corresponding to the specified state variable in the second state variable, divide multiple partitions according to the first value, the second value and the third value, and determine the comprehensive area model corresponding to the water pump and water turbine in each partition.

[0155] The control module 1205 is used to determine the parameter values ​​of the first state variable and the target value of the second state variable of the water pump turbine based on the integrated regional model, so as to control the water pump turbine according to the parameter values ​​and the target value.

[0156] In one exemplary embodiment of this disclosure, determining the predicted value of the second state variable based on the first state variable includes:

[0157] The first state variable and the second state variable are combined into a data matrix, and the first matrix and the second matrix are determined based on the data matrix;

[0158] By combining the observation function set, the first matrix and the second matrix are established as a linear relationship to generate a linear relationship between the first state variable and the second state variable;

[0159] Based on the initial values ​​and linear relationship of the first and second state variables, the predicted values ​​of the second state variables in other states are determined.

[0160] In one exemplary embodiment of this disclosure, the step of combining the observation function set to establish a linear relationship between the first matrix and the second matrix to generate a linear relationship between the first state variable and the second state variable includes:

[0161] Using the set of observation functions, a linear relationship is established between the first matrix and the second matrix, which is then transformed into an objective equation, which is a constant matrix.

[0162] The first matrix is ​​decomposed to obtain the component form. Based on the component form, the constant matrix is ​​transformed. The transformed constant matrix is ​​then decomposed by eigenvalues ​​to obtain the decomposed form.

[0163] The transformed constant matrix is ​​combined with the decomposition form to generate a linear relationship between the first state variable and the second state variable.

[0164] In one exemplary embodiment of this disclosure, the step of introducing an intermediate variable within the failure interval to determine a second value of the intermediate variable includes:

[0165] An intermediate variable is introduced within the failure interval, and a data matrix is ​​reconstructed based on the intermediate variable. The linear relationship between the intermediate variable and the second state variable is then determined again based on the reconstructed data matrix.

[0166] Based on the re-established linear relationship, the predicted value of the second state variable is determined, and the second value of the intermediate variable is determined based on the error between the predicted value and the actual value.

[0167] In one exemplary embodiment of this disclosure, determining the integrated region model corresponding to the pump turbine in each partition includes:

[0168] Within each partition, determine the partition fitting model for each second state variable composed of multiple first state variables;

[0169] By fitting different second-state variables to partitioned models and combining them according to the same partition, the comprehensive regional model corresponding to the water pump turbine in each partition is determined.

[0170] In one exemplary embodiment of this disclosure, the determination of the first value of the parameter corresponding to the error includes:

[0171] Gradually increase the value of the parameter until the calculated error is greater than the error threshold, and then determine the first value of the parameter.

[0172] In an exemplary embodiment of this disclosure, the first state variable includes the guide vane opening and unit rotational speed of the water pump turbine, the second state variable includes the flow rate and torque of the water pump turbine, and the intermediate variable is related to the target state variable and the second state variable.

[0173] It should be noted that the specific details of each module in the above-mentioned pump-turbine partition modeling device have been elaborated in detail in the corresponding pump-turbine partition modeling method, and will not be repeated here.

[0174] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0175] Furthermore, although the steps of the method in this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or a step may be broken down into multiple steps.

[0176] In an exemplary embodiment of this disclosure, an electronic device capable of implementing the above-described method is also provided.

[0177] Those skilled in the art will understand that various aspects of this disclosure can be implemented as a system, method, or program product. Therefore, various aspects of this disclosure can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software aspects, collectively referred to herein as a "circuit," "module," or "system."

[0178] The following reference Figure 13 To describe an electronic device 1300 according to such an embodiment of the present disclosure. Figure 13 The electronic device 1300 shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0179] like Figure 13As shown, the electronic device 1300 is manifested in the form of a general-purpose computing device. The components of the electronic device 1300 may include, but are not limited to: at least one processing unit 1310, at least one storage unit 1320, a bus 1330 connecting different system components (including storage unit 1320 and processing unit 1310), and a display unit 1340.

[0180] The storage unit stores program code that can be executed by the processing unit 1310, causing the processing unit 1310 to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure. For example, the processing unit 1310 can perform actions such as... Figure 1 The steps are shown in the figure.

[0181] Storage unit 1320 may include readable media in the form of volatile storage units, such as random access memory (RAM) 13201 and / or cache memory 13202, and may further include read-only memory (ROM) 13203.

[0182] Storage unit 1320 may also include a program / utility 13204 having a set (at least one) of program modules 13205, such program modules 13205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0183] Bus 1330 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the various bus structures.

[0184] Electronic device 1300 can also communicate with one or more external devices 1400 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 1300, and / or with any device that enables electronic device 1300 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 1350. Furthermore, electronic device 1300 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 1360. Figure 13As shown, network adapter 1360 communicates with other modules of electronic device 1300 via bus 1330. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with electronic device 1300, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0185] It should be noted that some embodiments of this disclosure also provide a computer program product, which includes a computer program that implements the above-described method when executed by a processor.

[0186] In one embodiment, the computer program product can be a tangible product containing a computer program, such as a computer-readable storage medium storing the computer program. The readable storage medium can be a storage medium based on electrical, magnetic, optical, electromagnetic, infrared, or other signals, including but not limited to: random access memory (RAM), read-only memory (ROM), magnetic tape, floppy disk, flash memory, hard disk drive (HDD), solid-state drive (SSD), etc. For example, the computer program product can be implemented as a non-volatile storage medium storing the computer program, such as read-only memory, NAND flash memory, etc. In one embodiment, the computer program product can be an intangible product containing a computer program. For example, the computer program product can be implemented as a virtual digital product, such as an executable file, installation package, or other digital file storing the computer program.

[0187] Computer program code can be written in one or more programming languages. Examples of programming languages ​​include C, Java, and C++. Program code can execute entirely on the user's computing device, partially on the user's computing device, or as a standalone software package. It can also execute partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, such as a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via an internet connection provided by a mobile network operator).

[0188] Computer programs can be carried or transmitted via signals such as electrical, magnetic, optical, electromagnetic, and infrared rays. Electronic devices can convert signals carrying computer programs into digital signals, thereby running the computer programs. When a computer program runs on an electronic device, its code is used to cause the electronic device to execute (more specifically, to be executed by the processor of the electronic device) the method steps of various exemplary embodiments of this disclosure.

[0189] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0190] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of this disclosure and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0191] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0192] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0193] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A method of partitioned modeling of a pump turbine, characterized by, For use in water pump turbines, the method includes: The first and second state variables in the experimental data of the water pump turbine are obtained, and the predicted value of the second state variable is determined based on the first state variable. The first state variable includes the guide vane opening and unit speed of the water pump turbine, and the second state variable includes the flow rate and torque of the water pump turbine. Determine the error between the actual value and the predicted value of the second state variable, and partition the target state variable in the first state variable according to the first value of the parameter corresponding to the error to determine multiple local intervals; If the error in any local interval is greater than the error threshold, an intermediate variable is introduced and a second value of the intermediate variable is determined; the intermediate variable is related to the unit speed, flow rate, and torque of the pump turbine. Determine the third value corresponding to the specified state variable in the second state variable, divide multiple partitions according to the first value, the second value and the third value, and determine the comprehensive region model corresponding to the water pump turbine in each partition; the specified state variable is the torque of the water pump turbine; Based on the comprehensive regional model, the parameter values ​​of the first state variable and the target value of the second state variable of the water pump turbine are determined, so as to control the water pump turbine according to the parameter values ​​and the target value; the control of the water pump turbine is to control the flow rate or the torque of the water pump turbine. The step of determining the predicted value of the second state variable based on the first state variable includes: The first state variable and the second state variable are combined into a data matrix, and the first matrix and the second matrix are determined based on the data matrix; By combining the observation function set, the first matrix and the second matrix are established as a linear relationship to generate a linear relationship between the first state variable and the second state variable; the observation function set consists of a first power observation function set and a second power observation function set, wherein the second power observation function set is selected according to the relationship between the flow rate of the water pump turbine and the guide vane opening, and the first power observation function set is selected according to the relationship between the flow rate of the water pump turbine and the unit speed. Based on the initial values ​​and linear relationship of the first and second state variables, the predicted values ​​of the second state variables in other states are determined.

2. The partition modeling method of a pump-turbine according to claim 1, characterized in that, The step of combining the observation function set to establish a linear relationship between the first and second matrices to generate a linear relationship between the first and second state variables includes: Using the set of observation functions, a linear relationship is established between the first matrix and the second matrix, which is then transformed into an objective equation, which is a constant matrix. The first matrix is ​​decomposed to obtain the component form. Based on the component form, the constant matrix is ​​transformed. The transformed constant matrix is ​​then decomposed by eigenvalues ​​to obtain the decomposed form. The transformed constant matrix is ​​combined with the decomposition form to generate a linear relationship between the first state variable and the second state variable.

3. The partition modeling method of a pump-turbine according to claim 1, characterized in that, The process of introducing an intermediate variable and determining its second value includes: An intermediate variable is introduced within the local interval, and a data matrix is ​​reconstructed based on the intermediate variable. The linear relationship between the intermediate variable and the second state variable is then determined again based on the reconstructed data matrix. Based on the re-established linear relationship, the predicted value of the second state variable is determined, and the second value of the intermediate variable is determined based on the error between the predicted value and the actual value.

4. The partition modeling method of a pump-turbine according to claim 1, characterized in that, The determination of the integrated regional model corresponding to the pump turbine in each partition includes: Within each partition, determine the partition fit model for each second state variable for multiple first state variables; The partitioned fitting models of multiple second-state variables are combined according to the same partition to determine the comprehensive regional model corresponding to the water pump turbine in each partition.

5. The partition modeling method of a pump-turbine according to claim 1, characterized by, The process of determining the first value of the parameter corresponding to the error includes: Gradually increase the value of the parameter until the calculated error is greater than the error threshold, and then determine the first value of the parameter.

6. A device for partition modeling of a pump turbine, characterized by include: The prediction value determination module is used to acquire the first state variable and the second state variable from the experimental data of the water pump turbine, and determine the predicted value of the second state variable based on the first state variable; the first state variable includes the guide vane opening degree and unit speed of the water pump turbine, and the second state variable includes the flow rate and torque of the water pump turbine. The initial partitioning module is used to determine the error between the actual value and the predicted value of the second state variable, and partition the target state variable in the first state variable according to the first value of the parameter corresponding to the error to determine multiple local intervals; The intermediate variable introduction module is used to introduce an intermediate variable and determine a second value of the intermediate variable if the error in any local interval is greater than the error threshold; the intermediate variable is related to the unit speed, flow rate and the torque of the water pump turbine; The model building module is used to determine the third value corresponding to the specified state variable in the second state variable, divide the system into multiple partitions based on the first value, the second value, and the third value, and determine the comprehensive region model corresponding to the water pump turbine in each partition; the specified state variable is the torque of the water pump turbine; The control module is used to determine the parameter values ​​of the first state variable and the target value of the second state variable of the water pump turbine based on the integrated regional model, so as to control the water pump turbine according to the parameter values ​​and the target value; the control of the water pump turbine is to control the flow rate or the torque of the water pump turbine. The step of determining the predicted value of the second state variable based on the first state variable includes: The first state variable and the second state variable are combined into a data matrix, and the first matrix and the second matrix are determined based on the data matrix; By combining the observation function set, the first matrix and the second matrix are established as a linear relationship to generate a linear relationship between the first state variable and the second state variable; the observation function set consists of a first power observation function set and a second power observation function set, wherein the second power observation function set is selected according to the relationship between the flow rate of the water pump turbine and the guide vane opening, and the first power observation function set is selected according to the relationship between the flow rate of the water pump turbine and the unit speed. Based on the initial values ​​and linear relationship of the first and second state variables, the predicted values ​​of the second state variables in other states are determined.

7. A computer program product, characterised in that, When the computer program is executed by the processor, it implements the partition modeling method for water pump turbines as described in any one of claims 1-5.

8. An electronic device, characterized in that, include: processor; Memory for storing the executable instructions of the processor; The processor is configured to implement the partition modeling method for a water pump turbine according to any one of claims 1-5 by executing the executable instructions.

Images