Method for evaluating wear life of ball valve sealing ring of hydroelectric generating set and application thereof

Abstract

The application provides a hydroelectric generating set ball valve sealing ring wear life evaluation method and application, relates to the intelligent operation and maintenance technical field of a hydropower station, and comprises the following steps: constructing a contact stress sample data set according to operation pressure data of a hydroelectric generating set ball valve sealing ring and corresponding contact stress reference values; constructing an initial neural network model and obtaining a target neural network model by training according to the contact stress sample data set; inputting real-time operation pressure data of the ball valve sealing ring into the target neural network model, outputting real-time equivalent contact stress, calculating the cumulative wear depth in a preset time according to the equivalent contact stress and a preset wear iteration model; calculating a real-time water seepage rate according to the cumulative wear depth and a gap flow theory, and predicting the remaining service life of the ball valve sealing ring according to the real-time water seepage rate. The application solves the problems of low simulation efficiency, discontinuous life evaluation and poor engineering operability of the existing sealing ring life evaluation technology.

Description

Technical Field

[0001] This application relates to the field of intelligent operation and maintenance technology for hydropower stations, and in particular to a method and application for assessing the wear life of ball valve sealing rings in hydropower units. Background Technology

[0002] The inlet ball valve of a high-head hydropower unit is a critical shut-off device in the power station's water transmission system. Failure of the ball valve's sealing ring can lead to pipeline leakage, reduced unit efficiency, and in severe cases, damage to pressure-bearing components. The sealing ring is subjected to alternating water pressure and reciprocating sliding friction over a long period, making wear and leakage a major cause of unit shutdown failures. Existing sealing ring life assessment technologies have the following significant shortcomings:

[0003] Low efficiency in obtaining contact stress: Contact stress in the sealing area is the core input for wear calculation. Traditional solutions rely on finite element simulation to solve for the contact stress in the sealing area. However, due to the real-time fluctuation of the unit's operating pressure, each set of pressure parameters needs to be calculated separately using finite element methods. Iterative simulation under multiple operating conditions is extremely time-consuming and cannot meet the needs of online real-time evaluation. Disconnection between wear and life assessment: Existing technologies can only calculate the instantaneous wear depth separately and have not established a coupling relationship between wear depth, sealing gap, water leakage rate, and failure threshold, making it impossible to continuously quantify the remaining service life. Lack of a unified quantitative safety rating standard, resulting in a lack of intuitive quantitative basis for maintenance decisions. Inability to achieve online calculation: Existing solutions have a broken data link and cannot directly output the sealing wear amount, water leakage rate, remaining service life, and safety status level from the unit's real-time operating pressure parameters with one click, making it difficult to adapt to on-site online monitoring systems.

[0004] Therefore, existing technologies suffer from problems such as low simulation efficiency, discontinuous life assessment, and poor engineering operability. There is an urgent need for a method to assess the wear life of ball valve seals that takes into account prediction accuracy, efficiency, and online real-time computing capabilities. Summary of the Invention

[0005] In view of this, this application proposes a method and application for assessing the wear life of ball valve seals in hydropower units, which can solve the problems of low simulation efficiency, discontinuous life assessment, and poor engineering operability of existing seal life assessment technologies.

[0006] In a first aspect, this application provides a method for assessing the wear life of ball valve seals in hydropower units, including: Obtain historical operating pressure data of the ball valve sealing ring of the target hydropower unit, and construct a contact stress sample dataset based on the operating pressure data and the contact stress benchmark value corresponding to the operating pressure data; An initial neural network model is constructed, and the initial neural network model is trained based on the contact stress sample dataset to obtain the target neural network model; The real-time operating pressure data of the ball valve sealing ring is input into the target neural network model, which outputs the real-time equivalent contact stress. The cumulative wear depth within a preset time is calculated based on the equivalent contact stress and the preset wear iteration model. The real-time seepage rate is calculated based on the cumulative wear depth and gap flow theory, and the remaining service life of the ball valve seal is predicted based on the real-time seepage rate.

[0007] In one embodiment, constructing the initial neural network model includes: By weighting the data-driven loss term and the physical prior loss term, a composite loss function for the initial neural network model is constructed. The data-driven loss term constrains the deviation between the initial neural network model prediction and the contact stress reference value, while the physical prior loss term constructs a proportional constraint between the operating pressure data and the initial neural network model prediction.

[0008] In one embodiment, the composite loss function is expressed as follows: In the formula, This represents the total loss of the model; For data-driven loss terms; This is the physical prior loss term; The weights of the physical prior loss term; The expression for the data-driven loss term is as follows: In the formula, Let be the network normalized prediction value of the i-th sample; This is the normalized baseline value; N The number of samples; The expression for the physical prior loss term is: In the formula, To average operating pressure, C This is the average scaling factor obtained from model training.

[0009] In one embodiment, calculating the cumulative wear depth over a preset time period based on the equivalent contact stress and a preset wear iteration model includes: The formula for calculating the wear depth of the ball valve seal ring during a single opening and closing cycle is as follows: In the formula, This refers to the depth of wear in a single instance. The wear coefficient; For sealing contact stress; This refers to the sliding distance during a single opening and closing operation. ; This refers to the travel length of the sealing ring. The hardness of the sealing ring material; The wear depth is accumulated daily to obtain the cumulative wear depth on day n: In the formula, The cumulative wear depth on day n; The wear depth is on day i.

[0010] In one embodiment, the step of calculating the real-time permeability based on the cumulative wear depth and gap flow theory includes: The formula for calculating real-time permeability is as follows: In the formula, Permeability; The equivalent diameter of the sealing contact area of ​​the ball valve; For fluid dynamic viscosity; This represents the cumulative wear depth. This refers to the operating pressure of the ball valve.

[0011] In one embodiment, predicting the remaining service life of the ball valve seal ring based on the real-time leakage rate includes: A preset critical seepage rate is defined, and the minimum cumulative wear depth at which the preset critical seepage rate is reached is obtained based on the relationship between the real-time seepage rate and the cumulative wear depth. The minimum total number of operating days for the ball valve seal ring is calculated based on the minimum cumulative wear depth. The remaining service life of the ball valve seal ring is then obtained based on the current number of operating days and the minimum total number of operating days.

[0012] In one embodiment, the method for assessing the wear life of the ball valve seal ring of the hydropower unit further includes: A piecewise nonlinear safety scoring model is constructed based on the relationship between the real-time seepage rate and the preset critical seepage rate of the project. The sealing status level of the ball valve sealing ring is determined based on the corresponding safety score in the safety scoring model, which is used to realize real-time early warning and maintenance guidance.

[0013] Secondly, this application also provides a device for assessing the wear life of a ball valve seal ring in a hydroelectric generator set, comprising: The dataset construction module is used to obtain historical operating pressure data of the ball valve sealing ring of the target hydropower unit, and construct a contact stress sample dataset based on the operating pressure data and the contact stress benchmark value corresponding to the operating pressure data. The model training module is used to construct an initial neural network model and train the initial neural network model based on the contact stress sample dataset to obtain a target neural network model. The cumulative wear depth calculation module is used to input the real-time operating pressure data of the ball valve sealing ring into the target neural network model, output the real-time equivalent contact stress, and calculate the cumulative wear depth within a preset time according to the equivalent contact stress and the preset wear iteration model. The remaining service life prediction module is used to calculate the real-time leakage rate based on the cumulative wear depth and gap flow theory, and predict the remaining service life of the ball valve seal ring based on the real-time leakage rate.

[0014] Thirdly, this application also provides an electronic device, including a processor and a memory; the memory has a stored computer program, wherein the computer program, when executed by the processor, implements the method for assessing the wear life of the ball valve seal ring of a hydroelectric generator as described in the first aspect.

[0015] Fourthly, this application also provides a computer storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the method for assessing the wear life of the ball valve seal ring of a hydroelectric generator as described in the first aspect.

[0016] The method for assessing the wear life of ball valve seals in hydropower units disclosed in this application has at least the following advantages compared to related technologies: 1. By acquiring historical operating pressure data of the ball valve sealing ring of the hydropower unit and constructing a contact stress sample dataset based on the corresponding contact stress benchmark value, a target neural network model is trained to achieve instantaneous output of contact stress with only pressure parameters input, solving the problem of excessive time consumption in multi-condition iterative simulation, meeting the requirements of online real-time evaluation, and establishing a coupled mathematical model of contact stress, cumulative wear depth, seepage rate, and failure threshold, which can fully realize the automatic calculation from operating pressure to remaining life.

[0017] 2. By constructing a physically constrained neural network model and designing a composite loss function, we can simultaneously constrain the data fitting accuracy and the stress-stress variation law. Compared with models such as BP neural network, regression tree, and ridge regression, the prediction error is lower and the engineering credibility is higher. It can balance prediction accuracy and physical rationality.

[0018] 3. By constructing a piecewise nonlinear safety scoring model, the cumulative wear depth, real-time seepage rate, remaining life, safety score, and sealing status level can be continuously output, which can be connected to the power plant online monitoring system to realize real-time early warning and maintenance guidance. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the method for assessing the wear life of the ball valve seal ring in a hydroelectric generator set according to one embodiment of this application; Figure 2 This is a schematic diagram illustrating the specific execution logic of the method for assessing the wear life of the ball valve seal ring of a hydroelectric generator unit in one embodiment of this application; Figure 3 This is a schematic diagram of the structure of a physical constraint neural network model in one embodiment of this application; Figure 4 This is a schematic diagram showing the comparison curves of average pressure-contact stress mapping for different models in one embodiment of this application; Figure 5 This is a schematic diagram showing the changes in the cumulative wear depth and water seepage rate of the sealing ring with the number of operating days in one embodiment of this application; Figure 6 This is a schematic diagram of the structure of the hydroelectric generator ball valve seal wear life assessment device in one embodiment of this application. Detailed Implementation

[0021] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0022] In some embodiments, such as Figure 1 As shown, this application provides a method for assessing the wear life of a ball valve seal ring in a hydroelectric generator set, comprising the following steps S101 to S104.

[0023] Step S101: Obtain historical operating pressure data of the ball valve sealing ring of the target hydropower unit, and construct a contact stress sample dataset based on the operating pressure data and the contact stress benchmark value corresponding to the operating pressure data.

[0024] Step S102: Construct an initial neural network model and train the initial neural network model based on the contact stress sample dataset to obtain the target neural network model.

[0025] Step S103: Input the real-time operating pressure data of the ball valve sealing ring into the target neural network model, output the real-time equivalent contact stress, and calculate the cumulative wear depth within a preset time based on the equivalent contact stress and the preset wear iteration model.

[0026] Step S104: Calculate the real-time seepage rate based on the cumulative wear depth and gap flow theory, and predict the remaining service life of the ball valve seal ring based on the real-time seepage rate.

[0027] Specifically, before implementing the method for assessing the wear life of ball valve seals in hydropower units as described in this application, for example, the equivalent contact stress reference value of the ball valve seal under different operating pressure conditions can be obtained through finite element simulation. Figure 2 The diagram shown is a schematic representation of the specific execution logic of the hydroelectric generator ball valve seal wear life assessment method of this application. Figure 2 As shown, in step S101, historical operating pressure data of the target hydropower unit's ball valve sealing ring is obtained, for example, based on the annual unit operating pressure data, to construct an original sample dataset, with the average pressure as the input feature for each sample. Maximum pressure and minimum pressure The output label is set to the corresponding equivalent contact stress reference value of the sealing ring. For example, the sample dataset is divided into training set, validation set, and test set according to a ratio of 7:1.5:1.5 for subsequent training, validation, and testing of the neural network model.

[0028] It should be noted that all input pressure variables and output contact stress variables are uniformly normalized to the interval [-1, 1] to eliminate the interference of differences in the dimensions of different physical quantities on model training. The normalization formula is: ; In the formula, These are the normalized variable values; x The original variable value; and These are the minimum and maximum values ​​of the variable in the sample, respectively.

[0029] In step S102, as Figure 3 As shown, the initial neural network model is constructed using a two-layer fully connected feedforward network, with 3 neurons in the input layer, corresponding to the average pressure, maximum pressure, and minimum pressure, respectively; each of the two hidden layers has 40 neurons, and the activation function is Tanh; the output layer outputs a normalized equivalent contact stress prediction value by a single neuron.

[0030] For example, the Adam optimization algorithm is used for model training, with an initial learning rate of 0.001 and a total of 800 training epochs. The samples are randomly shuffled and backpropagation training is performed in mini-batch with a batch size of 64. The optimal network parameters are selected using the validation set loss to complete the initial neural network model training and obtain the target neural network model.

[0031] In step S103, the real-time operating pressure data of the ball valve sealing ring is input into the trained target neural network model, and the real-time equivalent contact stress of the sealing ring is output. σ In this way, by constructing a neural network model to predict the contact stress of the sealing ring, the contact stress can be output instantaneously by only inputting pressure parameters, avoiding a large number of repetitive finite element simulations, solving the problem of excessive time consumption in multi-condition iterative simulations, and meeting the needs of online real-time evaluation.

[0032] Next, based on this equivalent contact stress σ A wear iteration model is constructed based on wear-related parameters of the ball valve seal ring. The cumulative wear depth within a preset time is calculated using this model. In one embodiment, the formula for calculating the wear depth of the ball valve seal ring during a single opening and closing cycle is first constructed as follows:

[0033] ; In the formula, This refers to the depth of wear in a single instance. The wear coefficient; For sealing contact stress; This refers to the sliding distance during a single opening and closing operation. ; This refers to the travel length of the sealing ring. The hardness of the sealing ring material.

[0034] The wear depth is then accumulated daily to obtain the cumulative wear depth on day n: ; In the formula, The cumulative wear depth on day n; The wear depth is on day i.

[0035] In step S104, the real-time leakage rate of the ball valve sealing ring is calculated based on the gap flow theory. For example, under low-speed conditions, ignoring the influence of relative sealing motion, the leakage rate calculation expression is constructed as follows: ; In the formula, Permeability; The equivalent diameter of the ball valve sealing contact area; For fluid dynamic viscosity; This represents the cumulative wear depth. This refers to the operating pressure of the ball valve.

[0036] Furthermore, the permeability calculation model based on the low-velocity gap flow theory will include the cumulative wear depth. Equivalent to the equivalent gap of the sealing pair, the real-time permeability is obtained. Real-time permeability With cumulative wear depth It exhibits cubic nonlinear growth.

[0037] In this way, the hydropower unit ball valve seal wear life assessment method of this embodiment establishes a coupled mathematical model of contact stress, cumulative wear depth, seepage rate and failure threshold, which can fully realize the automatic calculation from operating pressure to remaining service life.

[0038] In some embodiments, the initial neural network model is a physically constrained neural network contact stress proxy model, and its composite loss function is composed of a weighted average of a data-driven loss term and a physical prior loss term. The data-driven loss term constrains the deviation between the predicted value of the initial neural network model and the contact stress baseline value, while the physical prior loss term constructs a proportional constraint between the operating pressure data and the predicted value of the initial neural network model.

[0039] Furthermore, the expression for the composite loss function is: ; In the formula, This represents the total loss of the model; For data-driven loss terms; This is the physical prior loss term; The weights of the physical prior loss term. The results of the validation set were determined through debugging.

[0040] The data-driven loss term is represented by mean squared error, and its expression is as follows: ; In the formula, Let be the network normalized prediction value of the i-th sample; This is the normalized baseline value; N This represents the number of samples.

[0041] Physical prior loss term A proportional constraint between operating pressure and contact stress is constructed to ensure that the prediction results conform to the laws of sealing mechanics. Its expression is: ; In the formula, To average operating pressure, C This refers to the average scaling factor obtained during model training. The formula for calculating the average scaling factor is: .

[0042] It should be noted that conventional machine learning models suffer from poor physical consistency in prediction results. Data-driven models such as BP neural networks, regression trees, and ridge regression rely solely on sample fitting mapping relationships, lacking constraints from physical laws. This can lead to localized abnormal fluctuations in predicted contact stress, violating the sealing mechanics principle that "increased pressure leads to a synchronous increase in contact stress," resulting in low engineering reliability. In contrast, the neural network contact stress surrogate model in this embodiment uses a composite loss function that simultaneously constrains both data fitting accuracy and the pressure-stress variation law. Compared to BP neural networks, regression trees, and ridge regression models, it exhibits lower prediction errors and higher engineering reliability, balancing prediction accuracy with physical plausibility.

[0043] In some embodiments, such as Figure 5 As shown, for example, by setting a critical permeability rate for the project. q c As a criterion for seal failure, based on real-time water leakage rate With cumulative wear depth The relation calculation is satisfied ≥ q c The minimum total number of operating days is the total lifespan of the seal ring in case of failure. Combined with the current number of days the ball valve seal has been in operation. The remaining service life of the ball valve seal ring is calculated as follows: .

[0044] In some embodiments, after predicting the remaining service life of the ball valve seal, a piecewise nonlinear safety scoring model is constructed based on the relationship between the real-time leakage rate and the preset critical leakage rate, defining the leakage rate ratio as: .

[0045] For example, in the constructed piecewise nonlinear safety scoring model, the safety score... The scores range from 0 to 100, and the operating status is divided into 5 levels based on the scores, as shown in Table 1.

[0046] Table 1 Safety Score and Status Level Classification

[0047] In this way, by taking the real-time operating pressure of the ball valve sealing ring as input and iteratively running the above steps, the cumulative wear depth, real-time leakage rate, remaining life, safety score, and sealing status level can be continuously output, which can be connected to the power plant online monitoring system to realize real-time early warning and maintenance guidance.

[0048] The technical solution and effects of the method for assessing the wear life of a ball valve sealing ring in a hydropower unit provided in this application are further explained below with reference to specific embodiments. The implementation object is a ball valve sealing ring of a high-head hydropower station, and the data and basic parameter configuration are as follows: (1) Sample dataset: Collect the finite element simulation results of ball valves and construct 365 sets of annual operation samples with pressure range of 3.5~4.0MPa and sealing contact stress range of 90~95MPa; divide the training set into 255 sets, the validation set into 55 sets, and the test set into 7:1.5:1.5. (2) Fixed parameters of materials and working conditions: wear coefficient K = 5.0 × 10-7, material hardness H = 170 MPa, critical permeability =2.550 mL / s, sealing ring stroke length L=2280mm, single opening and closing sliding distance S=2L=4560mm; (3) Model hyperparameters: 40 hidden layer neurons, Adam initial learning rate 0.001, 800 training rounds, and activation function set to Tanh.

[0049] Complete the training and sample testing of the physical constraint neural network model, and combine Figure 4 The comparison curves of average pressure-contact stress mapping for different models are shown. Test results indicate that the RMSE of this embodiment is 0.151 MPa, MAE is 0.111 MPa, R² is 0.976, and MAPE is 0.122%. The residuals are concentrated near zero, and the prediction curve closely follows the physical law of increasing pressure-stress, without local abrupt changes, demonstrating superior overall performance.

[0050] Input the stress data for all 365 days of the year into the model for iterative calculation: (1) such as Figure 5 As shown, after 365 days of operation, the cumulative wear depth was 0.14 mm, and the real-time seepage rate was 0.403 mL / s, which was only 15.804% of the critical seepage rate. (2) Substituting into the safety scoring formula, we get The sealing condition is rated as good. (3) Continue iterating until the water seepage rate reaches the critical threshold. The total failure life is calculated to be 2305 days, and the remaining life of the current node is 1939 days.

[0051] In some embodiments, such as Figure 6 As shown in the figure, this application embodiment also provides a hydropower unit ball valve seal wear life assessment device 50, which includes: a dataset construction module 51, a model training module 52, a cumulative wear depth calculation module 53, and a remaining service life prediction module 54.

[0052] The dataset construction module 51 is used to obtain historical operating pressure data of the ball valve sealing ring of the target hydropower unit, and construct a contact stress sample dataset based on the operating pressure data and the contact stress benchmark value corresponding to the operating pressure data.

[0053] The model training module 52 is used to construct an initial neural network model and train the initial neural network model based on the contact stress sample dataset to obtain a target neural network model.

[0054] The cumulative wear depth calculation module 53 is used to input the real-time operating pressure data of the ball valve sealing ring into the target neural network model, output the real-time equivalent contact stress, and calculate the cumulative wear depth within a preset time according to the equivalent contact stress and the preset wear iteration model.

[0055] The remaining service life prediction module 54 is used to calculate the real-time leakage rate based on the cumulative wear depth and gap flow theory, and predict the remaining service life of the ball valve seal ring based on the real-time leakage rate.

[0056] It should be noted that the hydropower unit ball valve seal wear life assessment device 50 provided in this application embodiment and the water conveyance unit ball valve seal wear life assessment method provided in this application embodiment are based on the same inventive concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned hydropower unit ball valve seal wear life assessment method, and the repeated parts will not be described again.

[0057] In some embodiments, an electronic device provided in this application includes a processor and a memory; the memory stores a computer program, wherein the computer program, when executed by the processor, implements the above-described method for evaluating the wear life of the ball valve seal of a hydroelectric generator set.

[0058] Specifically, the processor may include, for example, a general-purpose microprocessor, an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor may also include onboard memory for caching purposes. The processor may be a single processing unit or multiple processing units for performing different actions of the method flow according to embodiments of this application.

[0059] Memory can be any medium capable of containing, storing, transmitting, propagating, or transmitting instructions. For example, memory can include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, instruments, or propagation media. Specific examples of memory include: magnetic storage devices such as magnetic tape or hard disk drives (HDDs); optical storage devices such as optical discs (CD-ROMs); and also random access memory (RAM) or flash memory; and / or wired / wireless communication links.

[0060] This application also provides a computer-readable medium storing a computer program that, when executed by a processor, implements the aforementioned method for assessing the wear life of a ball valve seal in a hydroelectric generator. This computer-readable medium may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into that device / apparatus / system. The aforementioned computer-readable medium carries one or more programs, which, when executed, implement the method as described in the embodiments of this application.

[0061] According to embodiments of this application, a computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wired, optical fiber, radio frequency signals, etc., or any suitable combination thereof.

[0062] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application. Therefore, the scope of this application should not be limited to the above embodiments. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for assessing the wear life of a ball valve seal in a hydroelectric generator set, characterized in that, include: Obtain historical operating pressure data of the ball valve sealing ring of the target hydropower unit, and construct a contact stress sample dataset based on the operating pressure data and the contact stress benchmark value corresponding to the operating pressure data; An initial neural network model is constructed, and the initial neural network model is trained based on the contact stress sample dataset to obtain the target neural network model; The real-time operating pressure data of the ball valve sealing ring is input into the target neural network model, which outputs the real-time equivalent contact stress. The cumulative wear depth within a preset time is calculated based on the equivalent contact stress and the preset wear iteration model. The real-time seepage rate is calculated based on the cumulative wear depth and gap flow theory, and the remaining service life of the ball valve seal is predicted based on the real-time seepage rate.

2. The method for assessing the wear life of the ball valve seal ring in a hydropower unit as described in claim 1, characterized in that, The construction of the initial neural network model includes: By weighting the data-driven loss term and the physical prior loss term, a composite loss function for the initial neural network model is constructed. The data-driven loss term constrains the deviation between the initial neural network model prediction and the contact stress reference value, while the physical prior loss term constructs a proportional constraint between the operating pressure data and the initial neural network model prediction.

3. The method for assessing the wear life of the ball valve seal ring in a hydropower unit as described in claim 2, characterized in that, The expression for the composite loss function is: ； In the formula, This represents the total loss of the model; For data-driven loss terms; This is the physical prior loss term; The weights of the physical prior loss term; The expression for the data-driven loss term is as follows: ； In the formula, Let be the network normalized prediction value of the i-th sample; This is the normalized baseline value; N The number of samples; The expression for the physical prior loss term is: ； In the formula, For average operating pressure, C This is the average scaling factor obtained from model training.

4. The method for assessing the wear life of the ball valve seal ring in a hydropower unit as described in claim 1, characterized in that, The step of calculating the cumulative wear depth within a preset time period based on the equivalent contact stress and a preset wear iteration model includes: The formula for calculating the wear depth of the ball valve seal ring during a single opening and closing cycle is as follows: ； In the formula, This refers to the depth of wear in a single instance. The wear coefficient; For sealing contact stress; This refers to the sliding distance during a single opening and closing operation. ; This refers to the travel length of the sealing ring. The hardness of the sealing ring material; The wear depth is accumulated daily to obtain the cumulative wear depth on day n: ； In the formula, The cumulative wear depth on day n; The wear depth is on day i.

5. The method for assessing the wear life of the ball valve seal ring in a hydropower unit as described in claim 1, characterized in that, The real-time permeability calculated based on the cumulative wear depth and gap flow theory includes: The formula for calculating real-time permeability is as follows: ； In the formula, Permeability; The equivalent diameter of the ball valve sealing contact area; For fluid dynamic viscosity; This represents the cumulative wear depth. This refers to the operating pressure of the ball valve.

6. The method for assessing the wear life of the ball valve seal ring in a hydropower unit as described in claim 1, characterized in that, The method of predicting the remaining service life of the ball valve seal ring based on the real-time leakage rate includes: A preset critical seepage rate is defined, and the minimum cumulative wear depth at which the preset critical seepage rate is reached is obtained based on the relationship between the real-time seepage rate and the cumulative wear depth. The minimum total number of operating days for the ball valve seal ring is calculated based on the minimum cumulative wear depth. The remaining service life of the ball valve seal ring is then obtained based on the current number of operating days and the minimum total number of operating days.

7. The method for assessing the wear life of the ball valve seal ring in a hydropower unit as described in claim 6, characterized in that, Also includes: A piecewise nonlinear safety scoring model is constructed based on the relationship between the real-time seepage rate and the preset critical seepage rate of the project. The sealing status level of the ball valve sealing ring is determined based on the corresponding safety score in the safety scoring model, which is used for real-time early warning and maintenance guidance.

8. A device for assessing the wear life of a ball valve seal in a hydroelectric generator set, characterized in that, include: The dataset construction module is used to obtain historical operating pressure data of the ball valve sealing ring of the target hydropower unit, and construct a contact stress sample dataset based on the operating pressure data and the contact stress benchmark value corresponding to the operating pressure data. The model training module is used to construct an initial neural network model and train the initial neural network model based on the contact stress sample dataset to obtain a target neural network model. The cumulative wear depth calculation module is used to input the real-time operating pressure data of the ball valve sealing ring into the target neural network model, output the real-time equivalent contact stress, and calculate the cumulative wear depth within a preset time according to the equivalent contact stress and the preset wear iteration model. The remaining service life prediction module is used to calculate the real-time leakage rate based on the cumulative wear depth and gap flow theory, and predict the remaining service life of the ball valve seal ring based on the real-time leakage rate.

9. An electronic device, characterized in that, It includes a processor and a memory; the memory stores a computer program, wherein the computer program, when executed by the processor, implements the method for assessing the wear life of the ball valve seal ring of a hydroelectric generator as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, It stores a computer program, wherein the computer program, when executed by a processor, implements the method for evaluating the wear life of the ball valve seal ring of a hydroelectric generator as described in any one of claims 1 to 7.

Images