A combined ventilation and water cooling method and device for a hydraulic power station

Abstract

This invention discloses a combined air-water cooling method and device for a hydraulic power station, relating to the field of hydraulic cooling control technology. The method includes: reading the expected die-casting plan of the target die-casting machine; determining the expected workload, predicting the temperature rise trend of the oil in multiple hydraulic components, and outputting multiple predicted oil temperatures; performing cooling demand analysis to determine multiple initial cooling temperature values; performing lubrication performance degradation analysis, compensating for the multiple initial cooling temperature values, and outputting multiple cooling temperatures; constructing a combined air-water cooling twin model, performing simulation optimization of cooling parameters, and outputting an optimized cooling control strategy; and performing cooling control on the target hydraulic power station during the cooling reflux cycle according to the optimized cooling control strategy. This invention solves the technical problems of insufficient cooling accuracy and practicality in existing technologies, which affect the working quality and efficiency of hydraulic power stations, achieving the technical effect of improving the cooling accuracy and practicality of hydraulic power stations, and enhancing their working efficiency and quality.

Description

Technical Field

[0001] This invention relates to the field of hydraulic heat dissipation control technology, specifically to a method and device for combined air and water heat dissipation in a hydraulic power station. Background Technology

[0002] In hydraulic power unit applications, especially in high-load, high-frequency operating environments such as die-casting machines, heat dissipation management of the hydraulic system is crucial. Existing heat dissipation control methods for hydraulic power units typically rely on combined air-cooling and water-cooling devices. These devices reduce the temperature of the hydraulic oil through fans and cooling water flow, ensuring that hydraulic components operate within a safe temperature range. However, existing technologies face some significant technical challenges in practical applications.

[0003] Existing methods for controlling the cooling of hydraulic stations using combined air-water cooling devices suffer from technical problems. They cannot accurately predict the temperature rise of the oil in each component of the hydraulic station based on the actual workload of the die-casting machine. This results in severe lag in the cooling control of each component, as well as insufficient accuracy and practicality in cooling, which affects the working quality and efficiency of the hydraulic station. Summary of the Invention

[0004] This application provides a combined air-water cooling method and device for a hydraulic power station, which addresses the technical problem that existing technologies cannot accurately predict the oil temperature rise of various components of the hydraulic power station based on the actual working load of the die-casting machine, resulting in insufficient accuracy and practicality in heat dissipation, which affects the working quality and efficiency of the hydraulic power station.

[0005] In view of the above problems, this application provides a method and device for combined air and water cooling of a hydraulic station.

[0006] The first aspect of this application provides a combined air-water cooling method for a hydraulic power station, the method comprising:

[0007] The process involves: reading the expected die-casting plan of the target die-casting machine, where the expected die-casting plan refers to the die-casting tasks within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station being the hydraulic component of the target die-casting machine; determining the expected workload based on the expected die-casting plan; predicting the temperature rise trend of the hydraulic fluid in multiple hydraulic components based on the expected workload, and outputting multiple predicted oil temperatures; performing a cooling demand analysis based on predetermined component oil temperature limits and the multiple predicted oil temperatures to determine multiple initial cooling temperature values; performing a lubrication performance degradation analysis based on the multiple predicted oil temperatures; compensating the multiple initial cooling temperature values ​​based on the multiple performance degradation values, and outputting multiple cooling temperatures; constructing a combined air-water cooling twin model based on the simulation of the target hydraulic station and the combined air-water cooling device; using the cooling twin model to optimize the cooling parameters by satisfying the multiple cooling temperature values ​​as cooling constraints and minimizing overall excessive heat dissipation, and outputting an optimized cooling control strategy; and implementing cooling control for the target hydraulic station within the cooling reflux cycle based on the optimized cooling control strategy.

[0008] A second aspect of this application provides a combined air-water cooling device for a hydraulic power station, the device comprising:

[0009] The system includes: a die-casting plan reading module, which reads the expected die-casting plan of the target die-casting machine, wherein the expected die-casting plan is the die-casting task within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station is the hydraulic component of the target die-casting machine; a plan parsing module, which determines the expected workload based on the expected die-casting plan, predicts the temperature rise trend of the hydraulic fluid in multiple hydraulic components based on the expected workload, and outputs multiple predicted oil temperatures; a cooling demand analysis module, which performs cooling demand analysis based on predetermined component oil temperature limits and the multiple predicted oil temperatures, and determines multiple initial cooling temperature values; and a temperature compensation module. The temperature compensation module performs lubrication performance degradation analysis based on the multiple predicted oil temperatures, compensates the multiple initial cooling temperature values ​​according to the multiple performance degradation values, and outputs multiple cooling temperatures; the simulation optimization module constructs a combined air-water cooling twin model based on the target hydraulic station and the combined air-water cooling device, uses the cooling twin model to perform cooling parameter simulation optimization with the goal of minimizing overall excessive heat dissipation, and outputs an optimized cooling control strategy; the cooling control module performs cooling control on the target hydraulic station during the cooling return cycle according to the optimized cooling control strategy.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] This application reads the expected die-casting plan of the target die-casting machine, where the expected die-casting plan is the die-casting task within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station is the hydraulic component of the target die-casting machine; based on the expected die-casting plan, the expected workload is determined; based on the expected workload, the temperature rise trend of the oil in multiple hydraulic components is predicted, and multiple predicted oil temperatures are output; based on the predetermined component oil temperature limit and multiple predicted oil temperatures, cooling demand analysis is performed to determine multiple initial cooling temperature values; based on the multiple predicted oil temperatures, lubrication performance degradation analysis is performed, and multiple initial cooling temperature values ​​are compensated based on multiple performance degradation values, and multiple cooling temperatures are output; based on the simulation of the target hydraulic station and the combined air-water cooling device, a combined air-water cooling twin model is constructed; using the cooling twin model, with multiple cooling temperature values ​​as cooling constraints, and with the aim of minimizing overall excessive heat dissipation, the cooling parameters are simulated and optimized, and an optimized cooling control strategy is output; based on the optimized cooling control strategy, the cooling control of the target hydraulic station is performed within the cooling reflux cycle. This invention addresses the technical problem of existing technologies failing to accurately predict the oil temperature rise of various components in a hydraulic station based on the actual workload of a die-casting machine. This results in insufficient accuracy and practicality in heat dissipation, impacting the working quality and efficiency of the hydraulic station. By reading the expected die-casting plan of the target die-casting machine, the working load of the hydraulic station is analyzed, thereby predicting the temperature rise trend of hydraulic components and determining the initial cooling temperature. Through lubrication performance degradation analysis, the initial cooling temperature is compensated, generating multiple cooling temperatures. A heat dissipation twin model is used for simulation and optimization, outputting an optimized heat dissipation control strategy, which is implemented during the cooling reflux cycle. This achieves the technical effect of improving the accuracy and practicality of heat dissipation in the hydraulic station, thereby enhancing its working efficiency and quality. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A schematic diagram of a combined air-water cooling method for a hydraulic station is provided in an embodiment of this application.

[0014] Figure 2 This is a schematic diagram of a combined air-water cooling device for a hydraulic station, provided as an embodiment of this application.

[0015] Figure labeling: Die casting plan reading module 11, plan parsing module 12, cooling demand analysis module 13, temperature value compensation module 14, simulation optimization module 15, heat dissipation control module 16. Detailed Implementation

[0016] This application provides a combined air-water cooling method and device for a hydraulic power station. It addresses the technical problem of existing technologies failing to accurately predict the oil temperature rise of various components in a hydraulic power station based on the actual workload of the die-casting machine, resulting in insufficient accuracy and practicality in heat dissipation, thus affecting the working quality and efficiency of the hydraulic power station. By reading the expected die-casting plan of the target die-casting machine, the working load of the hydraulic power station is analyzed, thereby predicting the temperature rise trend of hydraulic components and determining the initial cooling temperature. Through lubrication performance degradation analysis, the initial cooling temperature is compensated, generating multiple cooling temperatures. A heat dissipation twin model is used for simulation and optimization, outputting an optimized heat dissipation control strategy, which is implemented during the cooling reflux cycle. This achieves the technical effect of improving the accuracy and practicality of heat dissipation in the hydraulic power station, and enhancing its working efficiency and quality.

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0018] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, apparatus, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such process, method, product, or apparatus.

[0019] Example 1, as Figure 1 As shown, this application provides a combined air-water cooling method for a hydraulic power station, the method comprising:

[0020] Step S100: Read the expected die-casting plan of the target die-casting machine, wherein the expected die-casting plan is the die-casting task within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station is the hydraulic component of the target die-casting machine.

[0021] In this embodiment of the application, the expected die casting plan of the target die casting machine is obtained from the production management system of the target die casting machine. The expected die casting plan is a detailed arrangement of the production tasks of the die casting machine in the future working cycle, covering key contents such as the time nodes in the die casting process, the type of mold to be processed, the amount of material used, and the number of die castings.

[0022] The anticipated die-casting plan refers to the die-casting tasks within one cooling reflux cycle of the target hydraulic station. The cooling reflux cycle refers to the time period during which the circulating oil in the hydraulic station, after generating heat during the die-casting process, returns to the cooling device for cooling. The target hydraulic station is the hydraulic component of the target die-casting machine.

[0023] Step S200: Based on the expected die-casting plan, determine the expected workload, predict the temperature rise trend of the hydraulic fluid in multiple hydraulic components according to the expected workload, and output multiple predicted oil temperatures.

[0024] In this embodiment, relevant data, such as the start and end times of the task, required pressure, flow rate, and mold type, are first extracted from the anticipated die-casting plan. Then, a structured data parsing algorithm is used for formatting to extract key parameters related to the hydraulic station's workload. Next, a rule-based mapping algorithm is employed to associate the parsed task parameters with various hydraulic components of the hydraulic station. For example, high-pressure operation tasks primarily affect the load on the hydraulic pump, while frequent operation tasks primarily affect the load on the hydraulic valves. Through this task mapping, the specific load of each hydraulic component under different tasks is clearly defined.

[0025] Subsequently, a physics-based simulation algorithm was used to calculate the workload of each hydraulic component during the cooling recirculation cycle. For the hydraulic pump, its actual load under specific operating conditions was calculated using the flow-pressure relationship formula; for the hydraulic cylinder, its force and displacement under a specific pressure were calculated. After completing the workload calculation, an algorithm combining thermodynamic models and numerical simulation was used to predict the oil temperature rise trend. By simulating the heat conduction, convection, and radiation processes of the oil under different loads, curves showing the oil temperature change over time were generated, thereby predicting the oil temperature changes of each hydraulic component.

[0026] Finally, the temperature rise curves are summarized, and key indicators such as the peak and average temperatures of each hydraulic component during the cooling reflux cycle are extracted to generate multiple predicted oil temperatures.

[0027] Furthermore, the method provided in the application embodiments, which outputs multiple predicted oil temperatures, also includes:

[0028] Multiple hydraulic components of the target hydraulic station are acquired, and a first hydraulic component is randomly selected. Using the target die-casting machine, the target hydraulic station, and the first hydraulic component as feature association constraints, and guided by the oil temperature rise of the hydraulic component, information retrieval is performed using industrial big data to obtain a retrieval dataset. The feature similarity of the retrieval data meets a predetermined similarity threshold. The retrieval dataset is then subjected to difference dimensionality reduction to obtain a first sample dataset, which includes sample workload, sample initial oil temperature, and sample final oil temperature. Using the first sample dataset, a multi-source prediction operator is trained and cross-validated under supervision to obtain multiple convergent prediction units and multiple convergence degrees. Based on the principle of ensemble learning, a first prediction branch is generated by fusing the multiple convergent prediction units and the multiple convergence degrees. The first initial oil temperature of the first hydraulic component is acquired through a temperature sensing group. The expected workload and the first initial oil temperature are input into the first prediction branch, and the first predicted oil temperature is output and added to the multiple predicted oil temperatures.

[0029] In this embodiment, multiple hydraulic components of the target hydraulic station are first acquired, and one of them is randomly selected as the first hydraulic component. During this process, all hydraulic components, including hydraulic pumps, hydraulic motors, hydraulic cylinders, hydraulic valves, pipelines, and connectors, are identified and extracted from the hydraulic station's component library. Then, a random algorithm is used to select a hydraulic component and mark it as the first hydraulic component for subsequent analysis. Next, using the target die-casting machine, the target hydraulic station, and the first hydraulic component as feature association constraints, information retrieval is performed using an industrial big data platform.

[0030] In information retrieval, cosine similarity is used to evaluate the similarity between historical data and current operating conditions. Cosine similarity is a commonly used method that determines the similarity between two vectors by calculating the cosine of the angle between them. Specifically, the parameters of the current operating condition and historical data are first converted into feature vectors, and then the cosine similarity between them is calculated. The closer the similarity is to 1, the more similar the historical data is to the current operating condition. By matching task conditions, such as pressure, temperature, and load, the historical data most similar to the current operating condition is selected to form a retrieval dataset, where the feature similarity of the retrieved data meets a predetermined similarity threshold.

[0031] Subsequently, dimensionality reduction was performed on the retrieved dataset using techniques such as principal component analysis or t-SNE to reduce the data dimensionality while retaining key discriminative features, generating the first sample dataset. This dataset contains sample working loads, initial oil temperatures, and final oil temperatures related to the first hydraulic component. Next, using this first sample dataset, supervised training and cross-validation were performed on multi-source prediction operators (including Support Vector Machines (SVM), Backpropagation Neural Networks (BP Neural Networks), and Random Decision Forests). The trained model learned from the input sample data, and the accuracy and stability of the model were evaluated through cross-validation, ultimately yielding multiple convergent prediction units.

[0032] Convergence refers to the stability and accuracy of a model as it gradually approaches the optimal prediction result during training. To calculate the convergence of each convergent prediction unit, mean squared error or mean absolute error is typically used as an evaluation metric. These metrics measure the deviation between the predicted result and the true value. Higher convergence indicates higher prediction accuracy and less fluctuation. In multiple training and validation epochs, the error value trend of each model is recorded. If the error value tends to stabilize and the deviation is small, it indicates a high level of convergence.

[0033] Based on this, and using the principle of ensemble learning, the results of multiple convergent prediction units are weighted and fused to generate the first prediction branch. During weight confirmation, weights are allocated according to the convergence degree of each convergent prediction unit. Specifically, the higher the convergence degree of a model, the greater its weight in the final fusion. Weight allocation is achieved through normalization, that is, normalizing the convergence degree of all models. For example, the sum of all convergence degree values ​​is set to 1, and the normalized value is used as the weight of each model. During the fusion process, the prediction results of each model are weighted and summed according to their weights to form the final prediction output. Subsequently, the initial oil temperature of the first hydraulic component is obtained through a temperature sensor and input along with the expected workload into the first prediction branch model to output the first predicted oil temperature. Finally, the first predicted oil temperature is added to multiple predicted oil temperatures.

[0034] Furthermore, in the method provided in the application embodiments, obtaining the first sample dataset further includes:

[0035] First search data is randomly selected from the search dataset, wherein the first search data includes a first workload and a first initial oil temperature; deviation analysis is performed on the first workload and other workloads in the search dataset, and the number of data with a load deviation greater than a predetermined load threshold is counted and set as a first difference frequency; if the first difference frequency is greater than a predetermined frequency threshold, deviation analysis is performed on the first initial oil temperature and other initial oil temperatures in the search dataset to obtain a second difference frequency; if the second difference frequency is greater than the predetermined frequency threshold, the first search data is added to the first sample dataset.

[0036] In this embodiment, a record is first randomly selected from the retrieval dataset, referred to as the first retrieval data. This process uses a random number generation algorithm to generate a random index and extract the first workload and the first initial oil temperature from the corresponding record in the retrieval dataset. Next, a deviation analysis is performed on the first workload and the workloads of other records in the retrieval dataset. By calculating the difference, the difference between the first workload and each other workload is calculated sequentially to obtain each difference value. Then, the system performs frequency statistics on these differences to determine how many differences exceed a preset load threshold, and the number of times the threshold is exceeded is counted as the first difference frequency.

[0037] If the first difference frequency is greater than a predetermined frequency threshold, it indicates a significant difference between the first workload and other data. Further deviation analysis is then performed on the first initial oil temperature and the initial oil temperatures recorded with other data. This process is similar to the workload analysis: first, the difference between the first initial oil temperature and each other initial oil temperature is calculated; then, the percentage of these differences exceeding a preset oil temperature threshold is counted to determine the second difference frequency. If the second difference frequency also exceeds a predetermined threshold, it means that the first retrieved data also differs significantly from other data in terms of initial oil temperature.

[0038] After completing these two bias analyses, the results of the first and second difference frequencies are combined to determine whether to add the first retrieved data to the first sample dataset. If both difference frequencies exceed their respective thresholds, the first retrieved data is added to the sample dataset.

[0039] Step S300: Analyze the cooling requirements based on the predetermined component oil temperature limit and the multiple predicted oil temperatures to determine multiple initial cooling temperature values.

[0040] In this embodiment, the maximum oil temperature of each hydraulic component is first calculated and obtained using a prior oil temperature prediction model. These maximum values ​​represent the highest temperature that each component may reach under specific operating conditions. Meanwhile, the predetermined oil temperature limit for each hydraulic component is set by the manufacturer or engineer based on material properties and design requirements; this is a crucial upper temperature limit to ensure the safe operation of the hydraulic components.

[0041] Next, a threshold comparison algorithm is used to compare the maximum oil temperature of each hydraulic component with its predetermined oil temperature limit. If the maximum oil temperature is close to or exceeds the predetermined oil temperature limit, the component is marked as "requiring cooling," indicating that the component may exceed the safe temperature during actual operation and there is a risk of overheating.

[0042] For hydraulic components marked as requiring cooling, initial cooling temperatures are calculated using a cooling model. This model is based on the physical characteristics of the hydraulic system, including heat exchange processes such as conduction, convection, and radiation. First, Fourier's law of heat conduction is used to calculate the heat transfer of hydraulic oil within the component. Then, the convective heat transfer equation is used to analyze the heat exchange between the hydraulic oil and the pipe wall during flow. Finally, the Stefan-Boltzmann law is applied to calculate the radiative cooling effect on high-temperature components. The initial cooling temperature for each component is determined by combining the calculation results of these heat exchange mechanisms. Specifically, a weighted average is used to synthesize the results of each heat exchange mechanism to determine the final initial cooling temperature. In the weighted average method, weights are assigned by technical experts based on the relative importance and actual contribution of each cooling mechanism, resulting in a balanced cooling temperature.

[0043] Through the above process, multiple initial cooling temperature values ​​are determined.

[0044] Step S400: Perform lubrication performance degradation analysis based on the multiple predicted oil temperatures, compensate the multiple initial cooling temperature values ​​according to the multiple performance degradation values, and output multiple cooling temperatures.

[0045] Furthermore, the method provided in the application embodiment, which outputs multiple cooling temperatures, also includes:

[0046] A first hydraulic component is randomly selected, and its initial oil temperature, predicted oil temperature, and initial cooling temperature are obtained. A first temperature rise interval is constructed based on the initial oil temperature and the predicted oil temperature. A sample lubrication dataset is retrieved, constrained by the target hydraulic station, the first hydraulic component, and the first oil type. A first oil temperature-lubrication performance mapping is constructed based on the sample lubrication dataset, where the sample lubrication data includes sample oil temperature and sample lubrication characteristics. A first lubrication performance index of the first hydraulic component is obtained. The first temperature rise interval is filtered based on the first lubrication performance index and the first oil temperature-lubrication performance mapping to determine a first attenuation interval. Lubrication performance attenuation analysis is performed based on the first attenuation interval, and a first performance attenuation value is output. The first performance attenuation value is input into a predetermined attenuation-cooling correlation table to obtain a first cooling value. The first initial cooling temperature value is compensated based on the first cooling value, and a first cooling temperature is output and added to the plurality of cooling temperatures.

[0047] In this embodiment, a hydraulic component in the hydraulic station, referred to as the first hydraulic component, is first selected using a random number generation algorithm. Then, the first initial oil temperature and the first predicted oil temperature of this component are obtained through sensor or database queries. The initial oil temperature represents the current actual temperature, while the predicted oil temperature is based on historical data or simulation model analysis, reflecting possible future temperature changes. Simultaneously, the first initial cooling temperature value associated with this hydraulic component is obtained.

[0048] Next, based on the initial and predicted oil temperatures, a first temperature rise interval is generated using an interval construction algorithm. This interval describes the range of temperature changes that the hydraulic component may experience under specific operating conditions. Subsequently, using the target hydraulic station, the first hydraulic component, and the oil type as constraints, historical data under similar operating conditions is retrieved using database querying and big data analysis techniques to form a sample lubrication dataset. This dataset contains the temperature and corresponding lubrication characteristics of the sample oil, such as viscosity and oxidation resistance. Based on the retrieved sample lubrication dataset, a first oil temperature-lubrication performance mapping is constructed using regression analysis or interpolation methods. This mapping relationship demonstrates the quantitative relationship between oil temperature and lubrication performance. Then, a first lubrication performance index, such as the current viscosity value of the lubricating oil, is obtained and compared with the oil temperature-lubrication performance mapping using a conditional filtering algorithm. The temperature range within the first temperature rise interval related to this lubrication performance index is selected, thereby determining the first decay interval.

[0049] After determining the attenuation range, a lubrication performance attenuation analysis is performed, calculating the lubrication performance deviation values ​​within multiple temperature ranges, i.e., the extent of lubrication performance decline after the temperature exceeds the maximum oil temperature within these ranges. The average of all lubrication performance deviation values ​​outside the ranges is calculated to determine the first performance attenuation value. This attenuation value reflects the overall decline in lubrication performance after the temperature exceeds the safe range. This performance attenuation value is then input into an attenuation-temperature reduction correlation table, and a matching algorithm is used to obtain the corresponding first temperature reduction value, which is used to compensate for the initial cooling temperature value. The attenuation-temperature reduction correlation table can be constructed based on historical sample data analysis.

[0050] Finally, the initial cooling temperature value is compensated based on the first cooling value, that is, the initial cooling temperature value is added to the cooling value to obtain the compensated first cooling temperature, which is then added to multiple cooling temperatures.

[0051] Furthermore, in the method provided in the application embodiments, constructing the first oil temperature-lubrication performance mapping further includes:

[0052] The sample lubrication dataset is clustered based on a predetermined temperature step size to determine multiple oil temperature thresholds and multiple lubrication feature sets. A first oil temperature threshold and a first lubrication feature set are randomly selected, where the lubrication features include oil viscosity and oxidation stability. A first lubrication feature is randomly selected from the first lubrication feature set, and the first oil viscosity and first oxidation stability of the first lubrication feature are obtained. The deviation between the first oil viscosity and the viscosity of other oils in the first lubrication feature set is calculated, and the number of times the viscosity deviation is less than a predetermined viscosity threshold is counted and set as the first high frequency. If the first high frequency is greater than the predetermined frequency threshold, the deviation between the first oxidation stability and the oxidation stability of other oils in the first lubrication feature set is calculated to obtain the second high frequency. If the second high frequency is greater than the predetermined frequency threshold, the first lubrication feature is set as the first high-frequency lubrication feature, and the first high-frequency lubrication feature set is obtained sequentially. The first high-frequency lubrication feature set is mean-calculated and weighted fused to output the first lubrication performance coefficient. Based on the decision classification principle, the first oil temperature-lubrication performance mapping is constructed according to the mapping relationship between the first oil temperature threshold and the first lubrication performance coefficient.

[0053] In this embodiment, the sample data is first segmented and clustered according to a predetermined temperature step size. The predetermined temperature step size refers to the interval value used when dividing temperature ranges, typically determined based on accuracy requirements and the temperature sensitivity of the lubricating oil. For example, the temperature step size might be set to 0.5℃ or 1℃, dividing the sample data into multiple ranges. The K-means clustering algorithm is then used to cluster the data within each temperature range, thereby generating multiple oil temperature thresholds and their corresponding lubrication feature sets. These feature sets include key lubrication parameters within that temperature range, such as oil viscosity and oxidation stability.

[0054] After clustering, a first oil temperature threshold and its corresponding first lubrication feature set are randomly selected using a random number generation algorithm. Then, a first lubrication feature is randomly selected from this feature set, and its first oil viscosity and first oxidation stability are extracted. To further analyze these features, a deviation analysis is performed between the first oil viscosity and the viscosities of other oils in the feature set, calculating the difference between each pair of viscosity values. If these differences are less than a predetermined viscosity threshold, they are counted as high-frequency events, and the number of all high-frequency events is recorded, termed the first high frequency. If the first high frequency exceeds a predetermined frequency threshold, a similar deviation calculation is performed on the first oxidation stability to obtain a second high frequency.

[0055] After identifying the high-frequency lubrication characteristics, these characteristics are integrated into a first high-frequency lubrication characteristic set. This feature set is then processed using a weighted average method, where the weights are pre-set based on the importance of the characteristics. By calculating the mean of the lubrication characteristics within the feature set and weighting each characteristic according to the pre-set weights, a first lubrication performance coefficient is finally generated.

[0056] Finally, based on the principle of decision classification, specifically the decision tree classification algorithm, the first oil temperature threshold and the first lubrication performance coefficient are associated to construct a mapping relationship between oil temperature and lubrication performance, thus obtaining the first oil temperature-lubrication performance mapping. Specifically, the first oil temperature threshold is first used as the input variable of the decision tree model, and the first lubrication performance coefficient is used as the output target variable. The construction process of the decision tree relies on calculating the information gain or Gini index for each temperature threshold to determine the optimal split point. Information gain measures the degree to which the purity of the dataset is improved after splitting the dataset by a certain temperature threshold; while the Gini index measures the heterogeneity of the split dataset. The smaller the index, the clearer the classification of the dataset and the higher the purity. At each node in the decision tree construction, the information gain or Gini index of all currently available temperature thresholds is first calculated. Then, the temperature threshold that maximizes the information gain or minimizes the Gini index is selected as the optimal split point for that node. This split distinguishes the lubrication performance within different temperature ranges. Next, the above process is recursively repeated for each segmented subset to further segment the dataset until the dataset purity reaches a set standard, such as the Gini index approaching 0, or the information gain no longer significantly increasing. The segmentation process then stops. Finally, a tree-structured model is generated, where each leaf node corresponds to a specific lubrication performance coefficient, which describes the lubrication performance at a specific temperature threshold. This method yields the oil temperature-lubrication performance mapping.

[0057] Step S500: Based on the simulation of the target hydraulic station and the combined air-water cooling device, construct a combined air-water cooling twin model. Using the cooling twin model, with the multiple cooling temperature values ​​as cooling constraints, perform cooling parameter simulation optimization with the aim of minimizing overall excessive heat dissipation, and output an optimized cooling control strategy.

[0058] In this embodiment, a combined air-water cooling twin model is first constructed based on simulation data of the target hydraulic station and the combined air-water cooling device. After constructing the cooling twin model, several preset cooling temperature values ​​are used as cooling constraints, requiring all hydraulic components in the model to remain within these predetermined cooling temperature ranges. Simultaneously, excessive overall heat dissipation must be avoided, meaning that while ensuring the hydraulic components are at a safe temperature, energy waste or system efficiency reduction must be minimized.

[0059] To achieve this goal, heat dissipation parameters, including those for the water pump, bypass valve, and air cooling, are optimized using simulation models. The core of the optimization process is simulation-based optimization, which involves simulating different parameter combinations to find the optimal heat dissipation solution. Specifically, initial values ​​are first set for the water pump, bypass valve, and air cooling parameters, determined by technical experts according to design standards. Next, a series of simulation tests are run using a heat dissipation twin model. Each simulation adjusts these heat dissipation parameters and records the temperature changes of each hydraulic component within the hydraulic station. After each simulation, the overall heat dissipation effect of the hydraulic station is evaluated, with particular attention paid to whether the cooling temperature requirements of each component are met, and to checking for any excessive heat dissipation.

[0060] To further optimize, optimization algorithms, such as genetic algorithms or particle swarm optimization, are employed to continuously adjust parameters such as the water pump flow rate, the bypass valve opening, and the airflow speed of the air-cooling system to achieve the optimal heat dissipation effect. The goal of optimization is to find a combination of parameters that minimizes energy consumption and maximizes heat dissipation efficiency while meeting cooling requirements.

[0061] After multiple rounds of simulation and optimization, the final optimized heat dissipation control strategy was output.

[0062] Furthermore, in the method provided in the application embodiments, the output optimized heat dissipation control strategy further includes:

[0063] Obtain predetermined heat dissipation parameters and predetermined heat dissipation parameter thresholds, wherein the predetermined heat dissipation parameters include water pump parameters, bypass valve parameters, and air cooling parameters; based on the predetermined heat dissipation parameter thresholds, randomly generate multiple initial heat dissipation parameters that satisfy a uniform distribution; using the heat dissipation twin model, with the multiple cooling temperature values ​​as heat dissipation constraints, screen the multiple initial heat dissipation parameters to determine multiple qualified heat dissipation parameters; use the heat dissipation twin model and heat dissipation evaluation function to evaluate the heat dissipation fitness of the multiple qualified heat dissipation parameters and output multiple heat dissipation fitness values; use a predetermined optimization algorithm to optimize the heat dissipation parameters based on the heat dissipation twin model, heat dissipation evaluation function, and multiple heat dissipation fitness values, and output the optimized heat dissipation control strategy.

[0064] In this embodiment, predetermined heat dissipation parameters and predetermined thresholds are first obtained from the historical operating data of the hydraulic station. This data includes the operating records of the heat dissipation system of the hydraulic station under different working conditions, such as the water pump flow rate setting, the opening degree of the bypass valve, and the wind speed of the air-cooling system.

[0065] Next, based on predetermined heat dissipation parameter thresholds, a random number generation algorithm is used to randomly generate multiple initial heat dissipation parameter combinations that satisfy a uniform distribution. These initial heat dissipation parameters are then simulated using a heat dissipation twin model, with multiple cooling temperature values ​​as heat dissipation constraints, to filter the initial heat dissipation parameters. The heat dissipation twin model is used here to simulate excessive heat dissipation, ensuring that the system does not waste energy or reduce system efficiency due to excessive heat dissipation. Through screening, several qualified heat dissipation parameters are determined; these parameters meet cooling requirements while avoiding excessive heat dissipation.

[0066] Next, a heat dissipation evaluation function is used to evaluate the heat dissipation fitness of these qualified heat dissipation parameters. Heat dissipation fitness measures the effectiveness of each set of heat dissipation parameters, that is, whether the system can achieve optimal heat dissipation while meeting cooling requirements. Multiple heat dissipation fitness scores are output based on the evaluation.

[0067] Finally, a predetermined optimization algorithm is used to optimize the heat dissipation parameters based on the heat dissipation twin model, the heat dissipation evaluation function, and multiple heat dissipation fitnesss. Specifically, firstly, multiple generated initial heat dissipation parameters are set as initial solutions, and these initial solutions are arranged in descending order of heat dissipation fitness to generate an initial solution sequence. From the initial solution sequence, the first Q initial solutions are selected as head solutions, and the last M initial solutions are selected as tail solutions, where Q and M are the sum of the number of initial solutions, and M is an integer multiple of Q. Using the Q head solutions as centers, the M tail solutions are randomly clustered to obtain Q neighborhoods, with the same number of tail solutions in each neighborhood. Within each neighborhood, the tail solutions are mutated once according to a predetermined variation length, with the head solutions as the mutation direction, to obtain Q updated neighborhoods. If the mutated tail solutions do not meet the parameter nominal range, i.e., do not meet the predetermined heat dissipation parameter threshold, then this mutation is not performed. The mutated tail solutions are evaluated based on the heat dissipation evaluation function to identify the Q updated neighborhoods. If the fitness of a tail solution in a certain updated neighborhood is greater than the fitness of the head solution, then the tail solution replaces the head solution. This process is iterated until the predetermined number of optimization attempts is reached. In the final step, Q current neighborhoods are output, and the head solution in the optimal neighborhood is selected as the optimized heat dissipation control strategy. The optimal neighborhood is the one with the largest sum of heat dissipation fitness among the Q current neighborhoods.

[0068] Through the above process, an optimized heat dissipation control strategy is obtained.

[0069] Furthermore, in the method provided in the application embodiment, the expression of the heat dissipation evaluation function is:

[0070]

[0071] Where P represents heat dissipation adaptability, M is the number of hydraulic components, m represents any one of the M hydraulic components, and w mLet S be the influence degree of the m-th hydraulic component within the target hydraulic station, S be the simulated cooling temperature value of the m-th hydraulic component, and S1 be the cooling temperature value of the m-th hydraulic component.

[0072] In this embodiment, P represents heat dissipation adaptability, a comprehensive index used to evaluate the overall heat dissipation effect of the hydraulic station. A higher P value indicates a higher adaptability of the heat dissipation system under the current parameter settings. M represents the number of hydraulic components; the hydraulic station contains multiple hydraulic components, and M represents the total number of these components. m The influence of the m-th hydraulic component within the target hydraulic station is represented by this coefficient, which indicates the weight of the m-th hydraulic component's influence on the overall heat dissipation adaptability. This coefficient is preset by technical experts. S represents the simulated cooling temperature value of the m-th hydraulic component, calculated using a heat dissipation twin model. S1 represents the cooling temperature value of the m-th hydraulic component, which is the cooling temperature achieved by the component in actual operation, typically the set target cooling temperature.

[0073] The heat dissipation fitness is obtained by calculating using the heat dissipation evaluation function described above.

[0074] Step S600: Perform heat dissipation control on the target hydraulic station during the cooling reflux cycle according to the optimized heat dissipation control strategy.

[0075] In this embodiment, after formulating an optimized heat dissipation control strategy, it is directly applied to the heat dissipation control during the cooling recirculation cycle of the target hydraulic station. First, based on the water pump parameters, bypass valve parameters, and air cooling parameters set in the optimized strategy, each heat dissipation device of the hydraulic station is configured. During the cooling recirculation cycle, the optimized heat dissipation control strategy is followed; specifically, the water pump flow rate, bypass valve opening, and air cooling system wind speed are controlled according to the strategy-set parameters for each heat dissipation component of the hydraulic station. Each cooling circuit operates under fixed parameters, ensuring that the hydraulic oil is cooled to a preset cooling temperature after passing through the cooling device.

[0076] Through the above process, heat dissipation control during the cooling reflux cycle of the target hydraulic station is completed.

[0077] In summary, the embodiments of this application have at least the following technical effects:

[0078] This application reads the expected die-casting plan of the target die-casting machine, where the expected die-casting plan is the die-casting task within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station is the hydraulic component of the target die-casting machine; based on the expected die-casting plan, the expected workload is determined; based on the expected workload, the temperature rise trend of the oil in multiple hydraulic components is predicted, and multiple predicted oil temperatures are output; based on the predetermined component oil temperature limit and multiple predicted oil temperatures, cooling demand analysis is performed to determine multiple initial cooling temperature values; based on the multiple predicted oil temperatures, lubrication performance degradation analysis is performed, and multiple initial cooling temperature values ​​are compensated based on multiple performance degradation values, and multiple cooling temperatures are output; based on the simulation of the target hydraulic station and the combined air-water cooling device, a combined air-water cooling twin model is constructed; using the cooling twin model, with multiple cooling temperature values ​​as cooling constraints, and with the aim of minimizing overall excessive heat dissipation, the cooling parameters are simulated and optimized, and an optimized cooling control strategy is output; based on the optimized cooling control strategy, the cooling control of the target hydraulic station is performed within the cooling reflux cycle. This invention addresses the technical problem of existing technologies failing to accurately predict the oil temperature rise of various components in a hydraulic station based on the actual workload of a die-casting machine. This results in insufficient accuracy and practicality in heat dissipation, impacting the working quality and efficiency of the hydraulic station. By reading the expected die-casting plan of the target die-casting machine, the working load of the hydraulic station is analyzed, thereby predicting the temperature rise trend of hydraulic components and determining the initial cooling temperature. Through lubrication performance degradation analysis, the initial cooling temperature is compensated, generating multiple cooling temperatures. A heat dissipation twin model is used for simulation and optimization, outputting an optimized heat dissipation control strategy, which is implemented during the cooling reflux cycle. This achieves the technical effect of improving the accuracy and practicality of heat dissipation in the hydraulic station, thereby enhancing its working efficiency and quality.

[0079] Example 2, based on the same inventive concept as the combined air-water cooling method for a hydraulic station in the foregoing examples, such as... Figure 2 As shown, this application provides a combined air-water cooling device for a hydraulic power station. The device and method embodiments in this application are based on the same inventive concept. The device includes:

[0080] The system includes: a die-casting plan reading module 11, which reads the expected die-casting plan of the target die-casting machine, wherein the expected die-casting plan is the die-casting task within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station is the hydraulic component of the target die-casting machine; a plan analysis module 12, which determines the expected workload based on the expected die-casting plan, predicts the temperature rise trend of the oil in multiple hydraulic components based on the expected workload, and outputs multiple predicted oil temperatures; a cooling demand analysis module 13, which performs cooling demand analysis based on the predetermined component oil temperature limit and the multiple predicted oil temperatures, and determines multiple initial cooling temperature values; and a temperature value compensation module. 14. The temperature compensation module 14 performs lubrication performance degradation analysis based on the multiple predicted oil temperatures, compensates the multiple initial cooling temperature values ​​according to the multiple performance degradation values, and outputs multiple cooling temperatures; 15. The simulation optimization module 15 constructs a combined air-water cooling twin model based on the target hydraulic station and the combined air-water cooling device, uses the cooling twin model to perform cooling parameter simulation optimization with the goal of minimizing overall excessive heat dissipation, and outputs an optimized cooling control strategy; 16. The cooling control module 16 performs cooling control on the target hydraulic station during the cooling return cycle according to the optimized cooling control strategy.

[0081] Furthermore, the device is also used to perform the following functions:

[0082] Multiple hydraulic components of the target hydraulic station are acquired, and a first hydraulic component is randomly selected. Using the target die-casting machine, the target hydraulic station, and the first hydraulic component as feature association constraints, and guided by the oil temperature rise of the hydraulic component, information retrieval is performed using industrial big data to obtain a retrieval dataset. The feature similarity of the retrieval data meets a predetermined similarity threshold. The retrieval dataset is then subjected to difference dimensionality reduction to obtain a first sample dataset, which includes sample workload, sample initial oil temperature, and sample final oil temperature. Using the first sample dataset, a multi-source prediction operator is trained and cross-validated under supervision to obtain multiple convergent prediction units and multiple convergence degrees. Based on the principle of ensemble learning, a first prediction branch is generated by fusing the multiple convergent prediction units and the multiple convergence degrees. The first initial oil temperature of the first hydraulic component is acquired through a temperature sensing group. The expected workload and the first initial oil temperature are input into the first prediction branch, and the first predicted oil temperature is output and added to the multiple predicted oil temperatures.

[0083] Furthermore, the device is also used to perform the following functions:

[0084] First search data is randomly selected from the search dataset, wherein the first search data includes a first workload and a first initial oil temperature; deviation analysis is performed on the first workload and other workloads in the search dataset, and the number of data with a load deviation greater than a predetermined load threshold is counted and set as a first difference frequency; if the first difference frequency is greater than a predetermined frequency threshold, deviation analysis is performed on the first initial oil temperature and other initial oil temperatures in the search dataset to obtain a second difference frequency; if the second difference frequency is greater than the predetermined frequency threshold, the first search data is added to the first sample dataset.

[0085] Furthermore, the device is also used to perform the following functions:

[0086] A first hydraulic component is randomly selected, and its initial oil temperature, predicted oil temperature, and initial cooling temperature are obtained. A first temperature rise interval is constructed based on the initial oil temperature and the predicted oil temperature. A sample lubrication dataset is retrieved, constrained by the target hydraulic station, the first hydraulic component, and the first oil type. A first oil temperature-lubrication performance mapping is constructed based on the sample lubrication dataset, where the sample lubrication data includes sample oil temperature and sample lubrication characteristics. A first lubrication performance index of the first hydraulic component is obtained. The first temperature rise interval is filtered based on the first lubrication performance index and the first oil temperature-lubrication performance mapping to determine a first attenuation interval. Lubrication performance attenuation analysis is performed based on the first attenuation interval, and a first performance attenuation value is output. The first performance attenuation value is input into a predetermined attenuation-cooling correlation table to obtain a first cooling value. The first initial cooling temperature value is compensated based on the first cooling value, and a first cooling temperature is output and added to the plurality of cooling temperatures.

[0087] Furthermore, the device is also used to perform the following functions:

[0088] The sample lubrication dataset is clustered based on a predetermined temperature step size to determine multiple oil temperature thresholds and multiple lubrication feature sets. A first oil temperature threshold and a first lubrication feature set are randomly selected, where the lubrication features include oil viscosity and oxidation stability. A first lubrication feature is randomly selected from the first lubrication feature set, and the first oil viscosity and first oxidation stability of the first lubrication feature are obtained. The deviation between the first oil viscosity and the viscosity of other oils in the first lubrication feature set is calculated, and the number of times the viscosity deviation is less than a predetermined viscosity threshold is counted and set as the first high frequency. If the first high frequency is greater than the predetermined frequency threshold, the deviation between the first oxidation stability and the oxidation stability of other oils in the first lubrication feature set is calculated to obtain the second high frequency. If the second high frequency is greater than the predetermined frequency threshold, the first lubrication feature is set as the first high-frequency lubrication feature, and the first high-frequency lubrication feature set is obtained sequentially. The first high-frequency lubrication feature set is mean-calculated and weighted fused to output the first lubrication performance coefficient. Based on the decision classification principle, the first oil temperature-lubrication performance mapping is constructed according to the mapping relationship between the first oil temperature threshold and the first lubrication performance coefficient.

[0089] Furthermore, the device is also used to perform the following functions:

[0090] Obtain predetermined heat dissipation parameters and predetermined heat dissipation parameter thresholds, wherein the predetermined heat dissipation parameters include water pump parameters, bypass valve parameters, and air cooling parameters; based on the predetermined heat dissipation parameter thresholds, randomly generate multiple initial heat dissipation parameters that satisfy a uniform distribution; using the heat dissipation twin model, with the multiple cooling temperature values ​​as heat dissipation constraints, screen the multiple initial heat dissipation parameters to determine multiple qualified heat dissipation parameters; use the heat dissipation twin model and heat dissipation evaluation function to evaluate the heat dissipation fitness of the multiple qualified heat dissipation parameters and output multiple heat dissipation fitness values; use a predetermined optimization algorithm to optimize the heat dissipation parameters based on the heat dissipation twin model, heat dissipation evaluation function, and multiple heat dissipation fitness values, and output the optimized heat dissipation control strategy.

[0091] Furthermore, the device is also used to perform the following functions:

[0092]

[0093] Where P represents heat dissipation adaptability, M is the number of hydraulic components, m represents any one of the M hydraulic components, and w m Let S be the influence degree of the m-th hydraulic component within the target hydraulic station, S be the simulated cooling temperature value of the m-th hydraulic component, and S1 be the cooling temperature value of the m-th hydraulic component.

[0094] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0095] The above description is only a preferred embodiment of this application and is not intended to limit this application. 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.

[0096] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A combined air-water cooling method for a hydraulic power station, characterized in that, include: Read the expected die-casting plan of the target die-casting machine, wherein the expected die-casting plan is the die-casting task within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station is the hydraulic component of the target die-casting machine; Based on the expected die-casting plan, the expected workload is determined, and the temperature rise trend of the hydraulic fluid in multiple hydraulic components is predicted according to the expected workload, and multiple predicted oil temperatures are output. Based on the predetermined component oil temperature limit and the multiple predicted oil temperatures, a cooling demand analysis is performed to determine multiple initial cooling temperature values. Based on the multiple predicted oil temperatures, a lubrication performance degradation analysis is performed, and the multiple initial cooling temperature values ​​are compensated according to the multiple performance degradation values ​​to output multiple cooling temperatures. Based on the simulation of the target hydraulic station and the combined wind and water cooling device, a combined wind and water cooling twin model is constructed. Using the cooling twin model, the cooling parameters are simulated and optimized to minimize overall excessive heat dissipation, while satisfying the multiple cooling temperatures as heat dissipation constraints. An optimized heat dissipation control strategy is then output. The heat dissipation control strategy described above is used to control the heat dissipation of the target hydraulic station during the cooling reflux cycle. It outputs multiple cooling temperatures, including: A first hydraulic component is randomly selected, and the first initial oil temperature, the first predicted oil temperature, and the first initial cooling temperature of the first hydraulic component are obtained. A first temperature rise range is constructed based on the first initial oil temperature and the first predicted oil temperature. Using the target hydraulic station, the first hydraulic component, and the first oil type as constraints, a sample lubrication dataset is retrieved and obtained. A first oil temperature-lubrication performance mapping is constructed based on the sample lubrication dataset, wherein the sample lubrication data includes sample oil temperature and sample lubrication characteristics. Obtain the first lubrication performance index of the first hydraulic component, and filter the first temperature rise range based on the first lubrication performance index and the first oil temperature-lubrication performance mapping to determine the first attenuation range; Based on the first attenuation range, a lubrication performance attenuation analysis is performed, and a first performance attenuation value is output. The first performance attenuation value is input into a predetermined attenuation-temperature reduction correlation table, and a first temperature reduction value is obtained by matching. The first initial cooling temperature value is compensated according to the first temperature reduction value, and a first cooling temperature is output and added to the plurality of cooling temperatures.

2. The combined air-water cooling method for a hydraulic station according to claim 1, characterized in that, Output multiple predicted oil temperatures, including: Multiple hydraulic components of the target hydraulic station are obtained, and a first hydraulic component is randomly selected. The target die-casting machine, the target hydraulic station, and the first hydraulic component are used as feature association constraints. Guided by the oil temperature rise of the hydraulic component, information retrieval is performed using industrial big data to obtain a retrieval dataset. The feature similarity of the retrieval data meets a predetermined similarity threshold. The retrieved dataset is subjected to differential dimensionality reduction to obtain a first sample dataset, wherein the first sample data includes sample workload, sample initial oil temperature and sample final oil temperature; Using the first sample dataset, supervised training and cross-validation of the multi-source prediction operator are performed to obtain multiple convergent prediction units and multiple convergence degrees. Based on the principle of ensemble learning, a first prediction branch is generated by fusing the multiple convergent prediction units and the multiple convergence degrees. The first initial oil temperature of the first hydraulic component is obtained by the temperature sensing group. The expected working load and the first initial oil temperature are input into the first prediction branch, and the first predicted oil temperature is output and added to the plurality of predicted oil temperatures.

3. The combined air-water cooling method for a hydraulic station according to claim 2, characterized in that, The first sample dataset is obtained, including: Randomly select first search data from the search dataset, wherein the first search data includes a first working load and a first initial oil temperature; Perform a deviation analysis on the first workload and other workloads in the retrieved dataset, and count the number of data whose load deviation is greater than a predetermined load threshold, which is set as the first difference frequency. If the first difference frequency is greater than a predetermined frequency threshold, then a deviation analysis is performed on the first initial oil temperature and other initial oil temperatures in the retrieved dataset to obtain a second difference frequency. If the second difference frequency is greater than the predetermined frequency threshold, then the first retrieved data is added to the first sample dataset.

4. The combined air-water cooling method for a hydraulic station according to claim 1, characterized in that, Constructing the first oil temperature-lubrication performance mapping includes: The sample lubrication dataset is clustered based on a predetermined temperature step size to determine multiple oil temperature thresholds and multiple lubrication feature sets. A first oil temperature threshold and a first lubrication feature set are randomly selected, wherein the lubrication features include oil viscosity and oxidation stability. Randomly select a first lubrication feature from the first lubrication feature set, and obtain the first oil viscosity and first oxidation stability of the first lubrication feature; The viscosity of the first oil and the viscosity of other oils in the first lubrication feature set are calculated to determine the deviation. The number of times the viscosity deviation is less than a predetermined viscosity threshold is counted and set as the first high frequency. If the first high frequency is greater than a predetermined frequency threshold, then the deviation between the first oxidation stability and other oxidation stability in the first lubrication feature set is calculated to obtain the second high frequency. If the second high frequency is greater than the predetermined frequency threshold, the first lubrication feature is set as the first high frequency lubrication feature, the first high frequency lubrication feature set is obtained in sequence, the mean of the first high frequency lubrication feature set is calculated and weighted fused, and the first lubrication performance coefficient is output. Based on the decision classification principle, the first oil temperature-lubrication performance mapping is constructed according to the mapping relationship between the first oil temperature threshold and the first lubrication performance coefficient.

5. The combined air-water cooling method for a hydraulic station according to claim 1, characterized in that, Output optimization heat dissipation control strategies include: Obtain predetermined heat dissipation parameters and predetermined heat dissipation parameter thresholds, wherein the predetermined heat dissipation parameters include water pump parameters, bypass valve parameters, and air cooling parameters; Based on the predetermined heat dissipation parameter threshold, multiple initial heat dissipation parameters that satisfy uniform distribution are randomly generated. Using the heat dissipation twin model, the multiple initial heat dissipation parameters are screened to determine multiple qualified heat dissipation parameters, with the multiple cooling temperatures as heat dissipation constraints. The heat dissipation twin model and heat dissipation evaluation function are used to evaluate the heat dissipation fitness of the multiple qualified heat dissipation parameters, and multiple heat dissipation fitness values ​​are output. Using a predetermined optimization algorithm, the heat dissipation parameters are optimized based on the heat dissipation twin model, heat dissipation evaluation function, and multiple heat dissipation fitnesss, and the optimized heat dissipation control strategy is output.

6. The combined air-water cooling method for a hydraulic station according to claim 5, characterized in that, The expression for the heat dissipation evaluation function is: ； in, Characterizing heat dissipation adaptability, For the number of hydraulic components, Characterization Any one of the hydraulic components For the first The influence of each hydraulic component within the target hydraulic station. For the first Simulated cooling temperature values ​​for each hydraulic component. For the first Cooling temperature values ​​for each hydraulic component.

7. A combined air-water cooling system for a hydraulic station, characterized in that, The device is used to perform a combined air-water cooling method for a hydraulic station as described in any one of claims 1-6, and the device comprises: The die casting plan reading module reads the expected die casting plan of the target die casting machine, wherein the expected die casting plan is the die casting task within one cooling reflux cycle of the target hydraulic station, and the target hydraulic station is the hydraulic component of the target die casting machine; The planning analysis module determines the expected workload based on the expected die-casting plan analysis, predicts the temperature rise trend of the oil in multiple hydraulic components based on the expected workload, and outputs multiple predicted oil temperatures. A cooling demand analysis module performs cooling demand analysis based on predetermined component oil temperature limits and multiple predicted oil temperatures to determine multiple initial cooling temperature values. A temperature value compensation module performs lubrication performance degradation analysis based on the multiple predicted oil temperatures, compensates the multiple initial cooling temperature values ​​according to the multiple performance degradation values, and outputs multiple cooling temperatures. The simulation optimization module is based on the target hydraulic station and the combined air-water cooling device to construct a combined air-water cooling twin model. Using the cooling twin model, the module performs simulation optimization of cooling parameters to minimize overall excessive heat dissipation, while satisfying the multiple cooling temperatures as heat dissipation constraints. The module outputs an optimized heat dissipation control strategy. A heat dissipation control module, which controls the heat dissipation of the target hydraulic station during the cooling reflux cycle according to the optimized heat dissipation control strategy.

Images