Predictive maintenance method and system for hydraulic equipment based on Bayesian classification model

Abstract

The present invention discloses a predictive maintenance method and system for hydraulic equipment based on a Bayesian classification model, and relates to the technical field of hydraulic equipment. The predictive maintenance method for hydraulic equipment based on a Bayesian classification model includes the following steps: obtaining operating state characteristic data; obtaining a load prediction model, using the load prediction model to predict the load state of the hydraulic equipment at a future moment, and obtaining the load data of the hydraulic equipment; identifying the factor data affecting the load of the hydraulic equipment; determining the key factors affecting the failure of the hydraulic equipment, and debugging and maintaining the hydraulic equipment. The present invention can effectively process the uncertainty and noise in the data through the Bayesian classification model, improve the prediction accuracy of the load state of the hydraulic equipment, and enable the maintenance team to more accurately foresee potential failures, thereby intervening in advance.

Description

Technical Field

[0001] The present invention relates to the technical field of hydraulic equipment, and in particular to a predictive maintenance method and system for hydraulic equipment based on a Bayesian classification model. Background Art

[0002] Hydraulic equipment has been widely used in many industries due to its high power density, long life and easy automation. With the continuous advancement of automation technology, the structure and system of these equipment have become increasingly complex and interdependent, which not only improves operational efficiency but also increases the challenge of maintenance. Once a failure occurs, it may cause damage to the equipment function at the least, and even cause the production line to suspend operation in severe cases, causing huge economic losses and environmental pollution. Therefore, regular maintenance and fault prevention become particularly important, which can not only improve the reliability and safety of equipment operation, but also significantly reduce operating costs. Effective maintenance strategies include regular inspections, the application of fault diagnosis technology, and predictive maintenance measures, which are key factors to ensure the continuous and stable operation of hydraulic systems. Through these comprehensive measures, the benefits of equipment can be maximized, its service life can be extended, and the safety and efficiency of the production environment can be ensured.

[0003] In the prior art, it is not easy to predict the load state of hydraulic equipment, so it is not easy for technicians to timely understand the upcoming needs and potential problems of the equipment, and it is not easy to quickly identify key factors affecting equipment performance, so that early intervention on these factors is impossible, which increases the probability of failure. It is also not easy to perform predictive maintenance on hydraulic equipment, which in turn increases the probability of sudden shutdown due to equipment failure, and also increases the additional cost incurred by emergency repair of equipment.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention proposes a predictive maintenance method and system for hydraulic equipment based on a Bayesian classification model, which solves the problem mentioned in the above background technology that the existing technology is not convenient for predicting the load state of hydraulic equipment, making it inconvenient for technicians to timely understand the upcoming needs and potential problems of the equipment, and is not convenient for quickly identifying key factors affecting equipment performance, making it impossible to intervene in these factors early, increasing the probability of failure, and making it difficult to perform predictive maintenance on hydraulic equipment, thereby increasing the probability of sudden shutdown due to equipment failure, and also increasing the problem of additional costs incurred by emergency repair of equipment.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0007] According to one aspect of the present invention, a method for predictive maintenance of hydraulic equipment based on a Bayesian classification model is provided. The method for predictive maintenance of hydraulic equipment based on a Bayesian classification model comprises the following steps:

[0008] S1. Acquire real-time operation status data of hydraulic equipment, and pre-process the real-time operation status data to obtain operation status characteristic data;

[0009] S2. Bring the operating status feature data into the Bayesian classification model for training to obtain a load prediction model, and use the load prediction model to predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment;

[0010] S3, analyzing the load data of the hydraulic equipment using a load analysis algorithm, and identifying the factor data that affects the load of the hydraulic equipment;

[0011] S4, comparing the identified data of factors affecting the load of the hydraulic equipment with the known data of factors affecting the failure of the hydraulic equipment, determining the key factors affecting the failure of the hydraulic equipment, and performing debugging and maintenance on the hydraulic equipment;

[0012] The operating state feature data is brought into the Bayesian classification model for training to obtain a load prediction model, and the load prediction model is used to predict the load state of the hydraulic equipment at the future moment. Obtaining the load data of the hydraulic equipment includes the following steps:

[0013] S21, obtaining running status characteristic data, and dividing the running status characteristic data into a training set and a test set;

[0014] S22, using an integrated learning model, taking a certain operating status feature data as a super parent, and jointly constructing multiple Bayesian classification models with other operating status feature data;

[0015] S23, selecting the best Bayesian classification model on the training set from all constructed Bayesian classification models, and integrating the selected best Bayesian classification models to form a final load prediction model;

[0016] S24. Optimize the parameters of the load prediction model using a parameter optimization algorithm to find the optimal parameter configuration, apply the optimal parameter configuration to the final load prediction model, and predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment.

[0017] Furthermore, the calculation formula of the integrated learning model is:

[0018] ;

[0019] In the formula, A Expressed as hydraulic equipment load status category;

[0020] B Represented as a collection of operating status characteristic data;

[0021] Represented as the first i running status characteristic data value;

[0022] Expressed as i The conditional probability that the hydraulic equipment is in a load state when the characteristic data value of the operating state is equal to 1;

[0023] Expressed as Value and load status A In the case of j The conditional probability of the characteristic data value of the operating state;

[0024] n Expressed as the total number of operating status characteristic data.

[0025] Further, the parameters of the load prediction model are optimized using a parameter optimization algorithm to find the optimal parameter configuration, the optimal parameter configuration is applied to the final load prediction model, and the load state of the hydraulic equipment at a future moment is predicted. Obtaining the load data of the hydraulic equipment includes the following steps:

[0026] S241, obtaining historical operation data of the hydraulic equipment, dividing it into a training set and a test set, and calculating an initial fitness function value;

[0027] S242, preset the upper and lower limits of the load prediction model parameters, determine the dimension of the parameter configuration, and preset the population size and the maximum number of iterations;

[0028] S243. Initialize the parameter configuration population of the load prediction model according to the population size and the dimension of the parameter configuration;

[0029] S244, generating a reverse solution using a learning method, and calculating fitness function values ​​of the initial population and the reverse population;

[0030] S245, selecting the individuals with the best fitness from the initial population and the reverse population as the final initial population, and continuously updating the position and value of the best fitness;

[0031] S246. Adjust the parameter configuration position of the load prediction model according to the position update formula;

[0032] S247, repeating steps S244 to S246 until the set maximum number of iterations is reached, and the optimization process ends;

[0033] S248. Apply the optimal parameter configuration obtained in the iterative process to the constructed load prediction model, predict the load state of the hydraulic equipment at a future moment, and output the load data of the hydraulic equipment.

[0034] Furthermore, the formula for generating the reverse solution using the learning method is:

[0035] ;

[0036] In the formula, Represented as the reverse population c The position vector of the parameter configuration;

[0037] Represented as the first c The position vector of the parameter configuration;

[0038] S Represented as a random number;

[0039] P It represents the maximum value of the parameter configuration position vector in the initial population;

[0040] H Represents the minimum value of the parameter configuration position vector in the initial population.

[0041] Furthermore, the position update formula is:

[0042] ;

[0043] In the formula, represents the position vector of the parameter configuration in the next iteration;

[0044] The position vector of the parameter configuration in the current iteration;

[0045] Q Represented as parameter configuration;

[0046] x It is expressed as the golden section number;

[0047] Represented as a random number;

[0048] Represented as a random number;

[0049] W Expressed as an adjustment factor.

[0050] Further, using a load analysis algorithm to analyze the load data of the hydraulic equipment and identifying the factors affecting the load of the hydraulic equipment includes the following steps:

[0051] S31, identifying the factor that has the greatest impact on the load of the hydraulic equipment from the load data of the hydraulic equipment through a load analysis algorithm;

[0052] S32, determine the index for analyzing the load efficiency of hydraulic equipment, and set the efficiency threshold and corresponding improvement technology;

[0053] S33, if the load efficiency caused by the factor with the greatest impact on the load of the hydraulic equipment is lower than a preset threshold, the factor with the greatest impact is adjusted using a set improvement technology;

[0054] S34, establishing an optimization model for influencing factors in the hydraulic equipment load environment, and determining the objective function and constraint conditions of the optimization model;

[0055] S35. Apply the load analysis algorithm to the optimization model to find the best solution that meets the constraints and use it as the factor data affecting the load of the hydraulic equipment.

[0056] Furthermore, the load analysis algorithm is applied to the optimization model to find the best solution that meets the constraints, and the factors affecting the load of the hydraulic equipment include the following steps:

[0057] S351, processing the identified factors affecting the load of the hydraulic equipment to ensure that the factors affecting the load of the hydraulic equipment meet the preset operating conditions, and obtaining the processed factor data;

[0058] S352, initializing the parameters of the load analysis algorithm and setting the initial hydraulic equipment load optimization improvement technology;

[0059] S353, evaluating the specific impact of each improvement technology on the load of the hydraulic equipment through the established objective function, and updating the status according to the evaluation results;

[0060] S354, optimize, evaluate and adjust the improvement technology of factors affecting the load of hydraulic equipment, and continuously update the status according to the preset operating conditions;

[0061] S355. If the load efficiency of the optimized hydraulic equipment reaches a preset threshold, the best improvement technology is determined as the factor data affecting the load of the hydraulic equipment. If the load efficiency of the optimized hydraulic equipment does not reach the preset threshold, the improvement technology is continued to be optimized and adjusted.

[0062] Furthermore, comparing the identified factors affecting the load of the hydraulic equipment with the known factors of hydraulic equipment failure, determining the key factors affecting the failure of the hydraulic equipment, and debugging and maintaining the hydraulic equipment include the following steps:

[0063] S41, calculating the importance score of each factor affecting the load of the hydraulic equipment on the load of the hydraulic equipment according to the objective function, and selecting the factor that has the greatest impact on the failure of the hydraulic equipment as the key factor;

[0064] S42, performing a global search, calculating the impact of key factors under different settings on the performance of the hydraulic equipment, and comparing with the current equipment configuration;

[0065] S43. After each round of optimization, evaluate the new settings of all key factors and select the best equipment configuration as a new starting point for further optimization;

[0066] S44, updating the environmental factor settings according to the operation feedback and efficiency data of the hydraulic equipment;

[0067] S45, determining whether the optimization termination condition is met, if so, confirming the optimal setting, if not, returning to step S41 to continue optimization;

[0068] S46. Use the optimized and confirmed settings as the final adjustment of the environmental and operating conditions of the hydraulic equipment to ensure that the equipment operates in the best condition.

[0069] Furthermore, the formula for calculating the importance score of each factor affecting the load of the hydraulic equipment on the load of the hydraulic equipment according to the objective function is:

[0070] ;

[0071] In the formula, K It is expressed as the objective function;

[0072] D It is expressed as an importance membership matrix;

[0073] R It is expressed as a collection of factors that affect the load of hydraulic equipment;

[0074] Expressed as factors affecting the load of hydraulic equipment z Key factors affecting hydraulic equipment failure h The importance of membership;

[0075] e Expressed as the total number of factors affecting the load on the hydraulic equipment;

[0076] l Expressed as the total number of key factors affecting hydraulic equipment failure;

[0077] d It is expressed as fuzzy coefficient;

[0078] Expressed as factors affecting the load of hydraulic equipmentz Key factors affecting hydraulic equipment failure h The distance of influence between them.

[0079] According to another aspect of the present invention, a hydraulic equipment predictive maintenance system based on a Bayesian classification model is also provided. The hydraulic equipment predictive maintenance system based on a Bayesian classification model comprises:

[0080] The data acquisition module is used to acquire the real-time operation status data of the hydraulic equipment and pre-process the real-time operation status data to obtain the operation status characteristic data;

[0081] Establish a model module for bringing the operating status feature data into the Bayesian classification model for training to obtain a load prediction model, and use the load prediction model to predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment;

[0082] A load analysis module is used to analyze the load data of the hydraulic equipment using a load analysis algorithm and identify the factor data that affects the load of the hydraulic equipment;

[0083] The factor analysis module is used to compare the identified factor data affecting the load of hydraulic equipment with the known hydraulic equipment failure factor data, determine the key factors affecting the failure of hydraulic equipment, and perform debugging and maintenance on the hydraulic equipment;

[0084] The data acquisition module is connected with the load analysis module by establishing a model module, and the load analysis module is connected with the factor analysis module.

[0085] The beneficial effects of the present invention are:

[0086] 1. The present invention can effectively handle the uncertainty and noise in the data through the Bayesian classification model, improve the prediction accuracy of the load status of hydraulic equipment, and enable the maintenance team to more accurately foresee potential failures and intervene in advance. By continuously monitoring and analyzing the operating status data of hydraulic equipment, abnormal patterns can be discovered in time to prevent failures from occurring, thereby avoiding greater maintenance costs and equipment downtime. By analyzing load data and failure factors, the maintenance team can develop more targeted maintenance plans, focus resources on equipment components that really need maintenance, and optimize the use of maintenance resources. Predictive maintenance reduces sudden shutdowns caused by equipment failures, thereby helping enterprises avoid economic losses caused by production interruptions, while also reducing the additional costs incurred by emergency equipment repairs.

[0087] 2. The present invention can significantly improve the accuracy of the model in predicting the load of hydraulic equipment by using an integrated learning model and a parameter optimization algorithm. By integrating multiple Bayesian classification models and optimizing their parameter settings, the accuracy and reliability of the prediction results are ensured. In the process of setting model parameters, by setting the upper and lower limits of the parameters and optimizing the parameters using a population algorithm, the model's adaptability to different working conditions is increased, making it more flexible to adapt to various operating environments and improving the scope of application of the prediction. Efficient parameter optimization and accurate load prediction make the preventive maintenance of the equipment more effective. The operating strategy of the hydraulic equipment can be adjusted and optimized in time according to the prediction results to prevent potential failures, thereby reducing downtime and maintenance costs.

[0088] 3. The present invention uses load analysis algorithms to accurately identify key factors that affect the load of hydraulic equipment, and can perform equipment maintenance and adjustments more targetedly. By establishing optimization models and objective functions, the specific impact of improved technologies is systematically evaluated, thereby achieving precise maintenance. During the maintenance process, the hydraulic equipment is ensured to operate in the best state through continuous evaluation, optimization and adjustment. This not only improves the operating efficiency of the equipment, but also extends the service life of the equipment. By presetting thresholds and timely adjusting key factors that affect the load, the incidence of equipment failures is effectively reduced.

[0089] 4. The present invention uses the objective function to score key factors, which can accurately identify and sort the factors that have the greatest impact on the load of hydraulic equipment, so as to give priority to these factors to minimize the failure rate, perform global search and compare the impact of key factors on equipment performance under different settings, which is helpful to find the optimal configuration. Through continuous optimization iterations and equipment configuration updates, the method can continuously adjust the influencing factors, adapt to changes in the external environment and internal wear, ensure that the equipment is in the best operating state for a long time, update the environmental factor settings according to the equipment's operating feedback and efficiency data, so that the maintenance plan can respond to the actual operating conditions, further improve the reliability and performance of the equipment, determine whether the optimization reaches the termination conditions, and solidify the settings as the standard operating conditions of the equipment after confirmation, which helps to maintain the stability and predictability of equipment performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0091] Figure 1 is a flow chart of a method for predictive maintenance of hydraulic equipment based on a Bayesian classification model according to an embodiment of the present invention;

[0092] Figure 2 It is a principle block diagram of a hydraulic equipment predictive maintenance system based on a Bayesian classification model according to an embodiment of the present invention.

[0093] In the figure:

[0094] 1. Data acquisition module; 2. Model building module; 3. Load analysis module; 4. Factor analysis module. DETAILED DESCRIPTION

[0095] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0096] In the description of the present invention, unless otherwise specified, the meaning of "plurality" is two or more. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0097] According to an embodiment of the present invention, a method and system for predictive maintenance of hydraulic equipment based on a Bayesian classification model are provided.

[0098] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the predictive maintenance method of hydraulic equipment based on the Bayesian classification model of an embodiment of the present invention, the predictive maintenance method of hydraulic equipment based on the Bayesian classification model includes the following steps:

[0099] S1. Acquire real-time operation status data of hydraulic equipment, and pre-process the real-time operation status data to obtain operation status characteristic data;

[0100] Specifically, real-time operation status data includes:

[0101] Pressure data: The working pressure of the hydraulic system, including input pressure and output pressure.

[0102] Temperature data: The temperature of the hydraulic fluid and system components, which is critical to preventing overheating and tracking system efficiency.

[0103] Flow data: The oil flow of the hydraulic system shows the flow rate and amount of oil in the system.

[0104] Fluid Contamination: The cleanliness of the hydraulic fluid, including the type and amount of contaminants, which can indicate potential wear or failure.

[0105] Motor status data: Current and voltage data of the hydraulic pump motor, as well as the operating frequency.

[0106] Vibration Data: Data about the vibration patterns of a device, which can be used to detect imbalance, misalignment, or other mechanical problems.

[0107] S2. Bring the operating status feature data into the Bayesian classification model for training to obtain a load prediction model, and use the load prediction model to predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment;

[0108] Specifically, the load state refers to the workload that the equipment is subjected to under specific operating conditions.

[0109] Specifically, the load data includes:

[0110] Load pressure: The pressure value that hydraulic equipment is subjected to at different operating stages.

[0111] Load flow: The flow rate of oil during operation of the equipment, which affects the response speed and efficiency of the equipment.

[0112] Power demand: The power required by the device under different load conditions.

[0113] Load cycle: The period of load variation experienced by a device over a period of time, including peak loads and low load periods.

[0114] Load efficiency: How efficiently a device operates under a specific load, usually involving the ratio of energy consumption to output.

[0115] S3, analyzing the load data of the hydraulic equipment using a load analysis algorithm, and identifying the factor data that affects the load of the hydraulic equipment;

[0116] Specifically, the factors affecting the load of hydraulic equipment include:

[0117] Ambient temperature: Temperature conditions can significantly affect the viscosity of the hydraulic oil, which in turn affects the pressure and flow of the system.

[0118] Fluid quality: Fluid cleanliness, viscosity and other chemical properties are critical to the performance of a hydraulic system.

[0119] Frequency of use: The frequency and duty cycle of equipment operation have a direct impact on its degree of wear and tear and maintenance requirements.

[0120] Maintenance History: The maintenance and repair history of the equipment, including records of parts replaced and adjustments made that affect the current load capacity of the equipment.

[0121] Operating habits: The operator's usage habits, such as starting, stopping, and load change operations, may affect the load response of the equipment.

[0122] Load change rate: The rate at which loads change and frequent load fluctuations can cause additional stress on hydraulic equipment.

[0123] System Pressure Settings: The preset pressure value and adjustment of the hydraulic system have a significant effect on system response and load capacity.

[0124] Mechanical wear status: The degree of wear of components, such as seals, valves and pumps, affects system efficiency and load performance.

[0125] S4, comparing the identified data of factors affecting the load of the hydraulic equipment with the known data of factors affecting the failure of the hydraulic equipment, determining the key factors affecting the failure of the hydraulic equipment, and performing debugging and maintenance on the hydraulic equipment;

[0126] Specifically, the key factors affecting hydraulic equipment failure include:

[0127] Oil contamination: Contaminants in the hydraulic oil, such as moisture, air and solid particles, can cause increased wear inside the hydraulic system and affect the system's sealing and efficiency.

[0128] Oil aging: As hydraulic oil ages over time, its chemical properties change, which may lead to decreased lubricity and increased corrosiveness.

[0129] Component wear: Wear of key components such as seals, pumps, and valves is a common source of failure. Failure of these components usually causes system pressure loss or leakage.

[0130] System overload: Operating pressure or load beyond the design standards can cause premature wear or damage to hydraulic components.

[0131] Design defects: Defects in the hydraulic system design, such as unreasonable piping layout and insufficient capacity, may lead to system inefficiency and failure.

[0132] Abnormal temperature: System operating temperature that is too high or too low may affect the performance of the hydraulic oil and thus affect the stability of the entire system.

[0133] Operational errors: Improper operations by operators, such as incorrect startup and shutdown procedures, or inappropriate load adjustments, may cause damage to the system.

[0134] The operating state feature data is brought into the Bayesian classification model for training to obtain a load prediction model, and the load prediction model is used to predict the load state of the hydraulic equipment at the future moment. Obtaining the load data of the hydraulic equipment includes the following steps:

[0135] S21, obtaining running status characteristic data, and dividing the running status characteristic data into a training set and a test set;

[0136] Specifically, the training set generally accounts for 70%-80%, and the test set accounts for 20%-30%.

[0137] S22, using an integrated learning model (i.e., AODE learning model), taking a certain operating state feature data as a super-parent, and jointly constructing multiple Bayesian classification models (i.e., SPODE model) with other operating state feature data;

[0138] Specifically, the AODE (Averaged One-Dependence Estimators) learning model is an integrated Bayesian classification method that improves the basic naive Bayesian classifier to improve the accuracy and efficiency of classification. The AODE model introduces the concept of super-parent, allowing each attribute to show dependence under the condition of a given super-parent attribute, thereby reducing the restrictions of the attribute conditional independence assumption in naive Bayes.

[0139] S23, selecting the best Bayesian classification model on the training set from all constructed Bayesian classification models, and integrating the selected best Bayesian classification model (i.e., SPODE model) to form a final load prediction model;

[0140] Specifically, an ensemble learning method is used to integrate several SPODE models with the best accuracy to construct a final load prediction model. Common ensemble learning methods include voting method, averaging method, weighted average method, etc.

[0141] S24. Optimize the parameters of the load prediction model using a parameter optimization algorithm to find the optimal parameter configuration, apply the optimal parameter configuration to the final load prediction model, and predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment.

[0142] It should be explained that, taking hydraulic equipment operating heavy machinery in a large industrial environment as an example, first, real-time operating status data, including pressure, temperature, flow, etc., are collected from the sensors and log systems of the hydraulic equipment. The raw data is preprocessed, such as filtering and denoising, and useful operating status features, such as average pressure, temperature fluctuation, etc., are extracted. The extracted feature data are randomly divided into a training set (70%) and a test set (30%) to ensure that the model can be effectively tested on unseen data. A key feature is selected as the super parent, such as "average pressure", because pressure is usually the main indicator of the load of the hydraulic system. Multiple Bayesian classification models (SPODE) are constructed based on the super parent feature and other features. Each Bayesian classification model (SPODE) predicts different load aspects, such as pressure load, temperature load, etc. Each SPODE model is evaluated on the training set, and their accuracy is checked using methods such as cross validation. The best performing Bayesian classification models (SPODE) are selected, and these models are integrated into a comprehensive load prediction model using ensemble learning techniques (such as weighted average method). The weight may be based on the prediction accuracy of each Bayesian classification model (SPODE), and a parameter optimization algorithm, such as a genetic algorithm, a whale algorithm, etc., is selected to adjust the key parameters in the integrated load prediction model (e.g., the weights of each model), and the found optimal parameter configuration is applied to the load prediction model, and then the load prediction model is used to predict the load state of the hydraulic equipment at future moments.

[0143] Preferably, the calculation formula of the integrated learning model is:

[0144] ;

[0145] In the formula, A Expressed as hydraulic equipment load status category;

[0146] B Represented as a collection of operating status characteristic data;

[0147] Represented as the first i running status characteristic data value;

[0148] Expressed as i The conditional probability that the hydraulic equipment is in a load state when the characteristic data value of the operating state is equal to 1;

[0149] Expressed as Value and load status A In the case of j The conditional probability of the characteristic data value of the operating state;

[0150] n Expressed as the total number of operating status characteristic data.

[0151] Preferably, the parameters of the load prediction model are optimized using a parameter optimization algorithm to find the optimal parameter configuration, the optimal parameter configuration is applied to the final load prediction model, and the load state of the hydraulic equipment at a future moment is predicted. Obtaining the load data of the hydraulic equipment includes the following steps:

[0152] S241, obtaining historical operation data of the hydraulic equipment, dividing it into a training set and a test set, and calculating an initial fitness function value;

[0153] S242, preset the upper and lower limits of the load prediction model parameters, determine the dimension of the parameter configuration, and preset the population size and the maximum number of iterations;

[0154] S243. Initialize the parameter configuration population of the load prediction model according to the population size and the dimension of the parameter configuration;

[0155] S244, using a learning method (i.e., an elite reverse learning strategy) to generate a reverse solution, and calculating fitness function values ​​of the initial population and the reverse population;

[0156] Specifically, the elite reverse learning strategy is a technique commonly used in optimization algorithms, especially in genetic algorithms and other evolutionary algorithms. The purpose of this strategy is to enhance the algorithm's exploration ability and use the information of the current solution to improve the efficiency of future searches. The elite reverse learning strategy relies on selecting the best performing solutions (i.e., elite solutions) from the current population and using these solutions as the basis for generating new solutions. The "reverse" part of the strategy involves reversing certain parameters of the elite solution, that is, reversing or negating the values ​​of these parameters to explore new areas in the solution space.

[0157] S245, selecting the individuals with the best fitness from the initial population and the reverse population as the final initial population, and continuously updating the position and value of the best fitness;

[0158] S246. Adjust the parameter configuration position of the load prediction model according to the position update formula;

[0159] Specifically, the position update formula is the core process of the Golden Sine Algorithm (Gold-SA), which is the update process of its solution. The Golden Sine Algorithm (Gold-SA) introduces the golden section number based on the relationship between the sine function and the unit circle to achieve a good balance between search and development.

[0160] S247, repeating steps S244 to S246 until the set maximum number of iterations is reached, and the optimization process ends;

[0161] S248. Apply the optimal parameter configuration obtained in the iterative process to the constructed load prediction model, predict the load state of the hydraulic equipment at a future moment, and output the load data of the hydraulic equipment.

[0162] It should be explained that historical operating data, including pressure, temperature, flow, etc., are collected from the sensors of the hydraulic system, and the collected data are divided into a training set and a test set, usually with a ratio of 70% training set and 30% test set. A fitness function is established to evaluate the effect of the load prediction model parameter configuration. Usually, the upper and lower limits of the load prediction model parameters are determined based on the prediction accuracy of the load prediction model on the test set, such as the range of the load prediction model weight and learning rate. A population size, such as 50 candidate solutions, and a maximum number of iterations, such as 100, are set. According to the preset parameter dimension and population size, the initial parameter configuration is randomly generated. Set the population, generate a reverse solution for each parameter configuration, that is, the relative position in the parameter space, calculate the fitness function value of the initial population and its reverse population, select the individual with the best fitness from the initial population and the reverse population as the new population, record the current optimal fitness value and the corresponding parameter configuration, adjust the parameter configuration to improve the fitness according to the position update formula, such as using gradient descent, golden sine algorithm (Gold-SA), etc., apply the optimal parameter configuration obtained in the iterative process to the load prediction model, and finally use the optimized load prediction model to predict the load state of the hydraulic equipment at future times, and output related load data.

[0163] Specifically, the parameter optimization algorithm is an improved whale optimization algorithm based on the hybrid strategy of the elite reverse learning strategy and the golden sine algorithm. The hybrid strategy improved whale optimization algorithm proposed in the present invention retains the encirclement predation method and the random prey search method of the traditional whale algorithm, and uses the golden sine algorithm to improve the spiral ascending predation method of the whale. The improved position update method, such as the position update formula, introduces the golden section number. In each iteration, the position information is exchanged with the prey (the current optimal solution), and the direction and distance of the individual's movement are controlled by parameters, so that the individual whale approaches the direction of the prey (the current optimal solution) faster, which accelerates the convergence speed of the algorithm and the accuracy of optimization.

[0164] Preferably, the formula for generating the reverse solution using the learning method is:

[0165] ;

[0166] In the formula, Represented as the reverse population c The position vector of the parameter configuration;

[0167] Represented as the first cThe position vector of the parameter configuration;

[0168] S Represented as a random number;

[0169] Specifically, S is a random number in [0, 1].

[0170] P It represents the maximum value of the parameter configuration position vector in the initial population;

[0171] H Represents the minimum value of the parameter configuration position vector in the initial population.

[0172] Preferably, the location update formula is:

[0173] ;

[0174] In the formula, represents the position vector of the parameter configuration in the next iteration;

[0175] The position vector of the parameter configuration in the current iteration;

[0176] Q Represented as parameter configuration;

[0177] x It is expressed as the golden section number;

[0178] Specifically, x is the golden ratio number, which is a value of The fixed value of is approximately equal to 0.618. This value is used as a scaling factor in the algorithm to help adjust the dynamic balance between exploration and exploitation in order to achieve good convergence of the algorithm.

[0179] Represented as a random number;

[0180] Specifically, A random number between [0, 2π].

[0181] Represented as a random number;

[0182] Specifically, Represented as a random number between [0, π].

[0183] W Expressed as an adjustment factor.

[0184] Specifically, W Represents an adjustment factor used to calculate the new parameter configuration position vector.

[0185] Preferably, analyzing the load data of the hydraulic equipment using a load analysis algorithm to identify the factor data affecting the load of the hydraulic equipment includes the following steps:

[0186] S31, identifying the factor that has the greatest impact on the load of the hydraulic equipment from the load data of the hydraulic equipment through a load analysis algorithm;

[0187] S32, determine the index for analyzing the load efficiency of hydraulic equipment, and set the efficiency threshold and corresponding improvement technology;

[0188] S33, if the load efficiency caused by the factor with the greatest impact on the load of the hydraulic equipment is lower than a preset threshold, the factor with the greatest impact is adjusted using a set improvement technology;

[0189] S34, establishing an optimization model for influencing factors in the hydraulic equipment load environment, and determining the objective function and constraint conditions of the optimization model;

[0190] S35. Apply the load analysis algorithm to the optimization model to find the best solution that meets the constraints and use it as the factor data affecting the load of the hydraulic equipment.

[0191] It needs to be explained that, taking hydraulic equipment operating heavy machinery in a large industrial environment as an example, load data including pressure, temperature, flow, etc. are collected from sensors of hydraulic equipment, and principal component analysis (PCA) and fish school algorithm are applied to identify factors with the greatest impact on equipment load from the data. Energy efficiency ratio, load response time, etc. are set as indicators for evaluating the load efficiency of hydraulic equipment. Efficiency thresholds are set according to equipment performance requirements and historical data to determine when improvement measures need to be taken. If it is found that certain key factors cause the load efficiency to be lower than the threshold, the specific factors that need to be improved are identified. According to the nature of the influencing factors, corresponding improvement techniques are adopted, such as adjusting the pressure setting of the hydraulic system, improving the oil quality, or replacing worn parts, etc. An optimization model is established based on the identified key factors. The model includes objective functions and constraints. The objective function may be to minimize energy consumption or maximize output efficiency. The constraints include the operating parameters and safety standards of the equipment. The load analysis algorithm is applied in the optimization model to find the optimal parameter configuration that meets the constraints. The optimal solution found by the algorithm is applied to the actual equipment operation to improve the load management of hydraulic equipment.

[0192] Specifically, the load analysis algorithm is the fish school algorithm, which is an optimization algorithm based on swarm intelligence. It is inspired by the foraging, clustering and following behaviors of fish in nature. The algorithm imitates the collective behavior of fish in finding food and avoiding danger.

[0193] Preferably, applying the load analysis algorithm to the optimization model to find the best solution that meets the constraints and as the factor data affecting the load of the hydraulic equipment includes the following steps:

[0194] S351, processing (whitening) the identified factors affecting the load of the hydraulic equipment to ensure that the factors affecting the load of the hydraulic equipment meet the preset operating conditions, and obtaining the processed factor data;

[0195] S352, initializing the parameters of the load analysis algorithm (i.e., initializing the parameters of the fish school algorithm), and setting the initial hydraulic equipment load optimization improvement technology (i.e., artificial fish);

[0196] S353, evaluating the specific impact of each improvement technology on the load of the hydraulic equipment through the established objective function, and updating the status according to the evaluation results;

[0197] S354, optimizing (i.e., foraging behavior in the fish school algorithm), evaluating and adjusting (i.e., clustering behavior in the fish school algorithm) the improved technology of factors affecting the load of the hydraulic equipment, and continuously updating the status according to the preset operating conditions (i.e., the tail-chasing behavior in the fish school algorithm);

[0198] S355. If the load efficiency of the optimized hydraulic equipment reaches a preset threshold, the best improvement technology is determined as the factor data affecting the load of the hydraulic equipment. If the load efficiency of the optimized hydraulic equipment does not reach the preset threshold, the improvement technology is continued to be optimized and adjusted.

[0199] It should be explained that, taking hydraulic equipment operating heavy machinery in a large industrial environment as an example, data including pressure, flow, temperature, etc. are collected from the sensors of the hydraulic system, and the collected data is whitened to enhance the effect of subsequent algorithms by standardizing the data and removing correlations. It is ensured that the data has zero mean and unit variance statistically, as well as independence between different variables. The parameters of the fish school algorithm, such as the number of artificial fish, step size, field of view, etc., are initialized. The initial improvement technology is set, that is, the "artificial fish" representing different operating strategies in the simulation environment. The objective function is based on preset efficiency indicators, such as energy efficiency ratio or response time optimization. The performance of various improvement technologies (artificial fish) under the objective function is evaluated, and the state and position of the artificial fish are updated according to the evaluation results. Each artificial fish searches for a local optimal solution based on its current position, that is, tries to find the operating conditions that can best improve the load efficiency. If the local optimal solutions found by individuals are similar, they will form a group and move towards a better solution together. The artificial fish will try to follow the companions who have currently found the optimal solution, so as to quickly reach the global optimum. The operating parameters of the hydraulic equipment are continuously optimized through the fish school algorithm until the load efficiency reaches or exceeds the preset threshold. Once the best improvement technology that meets the threshold is found, it will be calibrated as a key factor affecting the load of the hydraulic equipment. If the optimized efficiency does not reach the preset threshold, the improvement technology will continue to be adjusted and optimized.

[0200] Preferably, comparing the identified factor data affecting the load of the hydraulic equipment with the known hydraulic equipment failure factor data, determining the key factors affecting the failure of the hydraulic equipment, and debugging and maintaining the hydraulic equipment includes the following steps:

[0201] S41, calculating the importance score of each factor affecting the load of the hydraulic equipment on the load of the hydraulic equipment according to the objective function (i.e., the objective function in the fuzzy C-means clustering algorithm), and selecting the factor that has the greatest impact on the failure of the hydraulic equipment as the key factor;

[0202] S42, performing a global search, calculating the impact of key factors under different settings on the performance of the hydraulic equipment, and comparing with the current equipment configuration;

[0203] S43, after each round of optimization (via the wolfpack optimization algorithm), evaluate the new settings of all key factors and select the best device configuration as a new starting point for further optimization;

[0204] S44, updating the environmental factor settings according to the operation feedback and efficiency data of the hydraulic equipment;

[0205] S45, determining whether the optimization termination condition is met, if so, confirming the optimal setting, if not, returning to step S41 to continue optimization;

[0206] S46. Use the optimized and confirmed settings as the final adjustment of the environmental and operating conditions of the hydraulic equipment to ensure that the equipment operates in the best condition.

[0207] It should be explained that, taking hydraulic equipment operating heavy machinery in a large industrial environment as an example, the operating data of the hydraulic equipment, including parameters such as pressure, flow, and temperature, are collected and analyzed. Statistical methods, fuzzy C-means clustering algorithms, etc. are used to calculate the importance of each factor on the load of the hydraulic equipment and determine the most critical factors. Through simulation or experimental methods, the impact of changes in key factors on equipment performance is tested under different operating conditions. These test results are compared with the current equipment configuration to identify possible opportunities for performance improvement or failure reduction. Based on the test results, the settings of key factors are adjusted and their impact on equipment performance is re-evaluated. The best performing configuration is selected from the iterative test as a new starting point for further optimization. According to the actual operation feedback and efficiency data of the hydraulic equipment, the settings of the environment and operating conditions are continuously updated. Taking into account changes in environmental factors such as temperature and humidity, the corresponding equipment settings are adjusted to adapt to these changes. The termination conditions of the optimization process are set, such as reaching the preset efficiency improvement or failure rate reduction goals. Once the termination conditions are reached, the best equipment settings are confirmed and fixed, and the optimized settings are applied to the actual equipment operation to ensure that the hydraulic equipment operates in the best state. After the new settings are implemented, the operating status and performance of the equipment continue to be monitored to ensure the sustainability of the optimization effect.

[0208] Specifically, the hybrid recommendation algorithm that integrates the wolf pack algorithm and fuzzy clustering is used to compare the identified factors affecting the load of hydraulic equipment with the known hydraulic equipment failure factors to determine the key factors affecting the failure of hydraulic equipment. In the load management of hydraulic equipment, data is often sparse, that is, data in certain time periods or under certain conditions may be incomplete. First, the item-based collaborative filtering method is used to pre-process the hydraulic equipment data, fill in missing data, and ensure the integrity of the analysis. The processed data is clustered by the fuzzy C-means clustering algorithm to divide the operating status of the hydraulic equipment into different modes, each of which represents a specific set of environmental and operating conditions. In the clustering process, the contribution of each operating mode to the load is determined according to the membership of the cluster. The wolf pack algorithm is introduced to optimize the fuzzy clustering process, and its global search capability is used to accurately find the operating parameters that most significantly affect energy consumption. This step is critical because it finds the best parameter combination through global optimal search to improve the energy efficiency of hydraulic equipment.

[0209] Preferably, the formula for calculating the importance score of each factor affecting the load of the hydraulic equipment on the load of the hydraulic equipment according to the objective function is:

[0210] ;

[0211] In the formula, K It is expressed as the objective function;

[0212] D It is expressed as an importance membership matrix;

[0213] R It is expressed as a collection of factors that affect the load of hydraulic equipment;

[0214] Expressed as factors affecting the load of hydraulic equipment z Key factors affecting hydraulic equipment failure h The importance of membership;

[0215] e Expressed as the total number of factors affecting the load on the hydraulic equipment;

[0216] l Expressed as the total number of key factors affecting hydraulic equipment failure;

[0217] d It is expressed as fuzzy coefficient;

[0218] Expressed as factors affecting the load of hydraulic equipment z Key factors affecting hydraulic equipment failure h The distance of influence between them.

[0219] According to another embodiment of the present invention, Figure 2 As shown, a hydraulic equipment predictive maintenance system based on a Bayesian classification model is also provided, and the hydraulic equipment predictive maintenance system based on a Bayesian classification model includes:

[0220] The data acquisition module 1 is used to acquire the real-time operation status data of the hydraulic equipment and pre-process the real-time operation status data to obtain the operation status characteristic data;

[0221] Establishing model module 2, which is used to bring the operating state feature data into the Bayesian classification model for training, obtain the load prediction model, and use the load prediction model to predict the load state of the hydraulic equipment at a future moment, and obtain the load data of the hydraulic equipment;

[0222] The load analysis module 3 is used to analyze the load data of the hydraulic equipment using a load analysis algorithm and identify the factor data that affects the load of the hydraulic equipment;

[0223] The factor analysis module 4 is used to compare the identified factor data affecting the load of the hydraulic equipment with the known hydraulic equipment failure factor data, determine the key factors affecting the hydraulic equipment failure, and perform debugging and maintenance on the hydraulic equipment;

[0224] The data acquisition module 1 is connected with the load analysis module 3 through the model establishment module 2, and the load analysis module 3 is connected with the factor analysis module 4.

[0225] In summary, with the help of the above technical solutions of the present invention, the present invention can significantly improve the accuracy of the model in hydraulic equipment load prediction by using an integrated learning model and a parameter optimization algorithm. By integrating multiple Bayesian classification models and optimizing their parameter settings, the accuracy and reliability of the prediction results are ensured. In the process of setting model parameters, by setting the upper and lower limits of the parameters and optimizing the parameters using a population algorithm, the adaptability of the model to different working conditions is increased, making it more flexible to adapt to various operating environments and improving the applicable scope of the prediction. Efficient parameter optimization and accurate load prediction make the preventive maintenance of the equipment more effective, and the operation strategy of the hydraulic equipment can be adjusted and optimized in time according to the prediction results to prevent potential failures, thereby reducing downtime and maintenance costs. The present invention uses a load analysis algorithm to accurately identify the key factors that affect the load of hydraulic equipment, and can perform equipment maintenance and adjustment more targetedly. By establishing an optimization model and an objective function, the specific impact of the improved technology is systematically evaluated, thereby achieving precise maintenance. During the maintenance process, the hydraulic equipment is ensured to operate in the best state through continuous evaluation, optimization and adjustment. It not only improves the operating efficiency of the equipment, but also extends the service life of the equipment. By presetting thresholds and timely adjusting the key factors that affect the load, the incidence of equipment failures is effectively reduced. The present invention uses the objective function to score key factors, which can accurately identify and sort the factors that have the greatest impact on the load of hydraulic equipment, so as to give priority to these factors to minimize the failure rate, perform global search and compare the impact of key factors under different settings on equipment performance, which helps to find the optimal configuration. Through continuous optimization iterations and equipment configuration updates, the method can continuously adjust the influencing factors, adapt to the external environment and internal wear changes, ensure that the equipment is in the best operating state for a long time, update the environmental factor settings according to the equipment's operating feedback and efficiency data, so that the maintenance plan can respond to the actual operating conditions, further improve the reliability and performance of the equipment, judge whether the optimization reaches the termination conditions, and solidify the settings as the standard operating conditions of the equipment after confirmation, which helps to maintain the stability and predictability of equipment performance.

[0226] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A predictive maintenance method for hydraulic equipment based on a Bayesian classification model, characterized in that: The predictive maintenance method for hydraulic equipment based on the Bayesian classification model includes the following steps: S1. Acquire real-time operation status data of hydraulic equipment, and pre-process the real-time operation status data to obtain operation status characteristic data; S2. Bring the operating status feature data into the Bayesian classification model for training to obtain a load prediction model, and use the load prediction model to predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment; S3, analyzing the load data of the hydraulic equipment using a load analysis algorithm, and identifying the factor data that affects the load of the hydraulic equipment; S4, comparing the identified data of factors affecting the load of the hydraulic equipment with the known data of factors affecting the failure of the hydraulic equipment, determining the key factors affecting the failure of the hydraulic equipment, and performing debugging and maintenance on the hydraulic equipment; The operation state characteristic data is brought into the Bayesian classification model for training to obtain a load prediction model, and the load prediction model is used to predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment, including the following steps: S21, obtaining running status characteristic data, and dividing the running status characteristic data into a training set and a test set; S22, using an integrated learning model, taking a certain operating status feature data as a super parent, and jointly constructing multiple Bayesian classification models with other operating status feature data; S23, selecting the best Bayesian classification model on the training set from all constructed Bayesian classification models, and integrating the selected best Bayesian classification models to form a final load prediction model; S24, optimizing the parameters of the load prediction model using a parameter optimization algorithm, finding the optimal parameter configuration, applying the optimal parameter configuration to the final load prediction model, and predicting the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment; The calculation formula of the integrated learning model is: ； In the formula, A Expressed as hydraulic equipment load status category; B Represented as a collection of operating status characteristic data; Represented as the first i running status characteristic data value; Expressed as i The conditional probability that the hydraulic equipment is in a load state when the characteristic data value of the operating state is equal to 1; Expressed as Value and load status A In the case of j The conditional probability of the characteristic data value of the operating state; n It is expressed as the total number of operating status characteristic data; The method of analyzing the load data of the hydraulic equipment by using the load analysis algorithm to identify the factor data affecting the load of the hydraulic equipment includes the following steps: S31, identifying the factor that has the greatest impact on the load of the hydraulic equipment from the load data of the hydraulic equipment through a load analysis algorithm; S32, determine the index for analyzing the load efficiency of hydraulic equipment, and set the efficiency threshold and corresponding improvement technology; S33, if the load efficiency caused by the factor with the greatest impact on the load of the hydraulic equipment is lower than a preset threshold, the factor with the greatest impact is adjusted using a set improvement technology; S34, establishing an optimization model for influencing factors in the hydraulic equipment load environment, and determining the objective function and constraint conditions of the optimization model; S35. Apply the load analysis algorithm to the optimization model to find the best solution that meets the constraints and use it as the factor data affecting the load of the hydraulic equipment.

2. The predictive maintenance method for hydraulic equipment based on the Bayesian classification model according to claim 1 is characterized in that: The method of using a parameter optimization algorithm to optimize the parameters of the load prediction model, finding the optimal parameter configuration, applying the optimal parameter configuration to the final load prediction model, and predicting the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment includes the following steps: S241, obtaining historical operation data of the hydraulic equipment, dividing it into a training set and a test set, and calculating an initial fitness function value; S242, preset the upper and lower limits of the load prediction model parameters, determine the dimension of the parameter configuration, and preset the population size and the maximum number of iterations; S243. Initialize the parameter configuration population of the load prediction model according to the population size and the dimension of the parameter configuration; S244, generating a reverse solution using a learning method, and calculating fitness function values ​​of the initial population and the reverse population; S245, selecting the individuals with the best fitness from the initial population and the reverse population as the final initial population, and continuously updating the position and value of the best fitness; S246. Adjust the parameter configuration position of the load prediction model according to the position update formula; S247, repeating steps S244 to S246 until the set maximum number of iterations is reached, and the optimization process ends; S248. Apply the optimal parameter configuration obtained in the iterative process to the constructed load prediction model, predict the load state of the hydraulic equipment at a future moment, and output the load data of the hydraulic equipment.

3. The predictive maintenance method for hydraulic equipment based on the Bayesian classification model according to claim 2 is characterized in that: The formula for generating the reverse solution using the learning method is: ； In the formula, Represented as the reverse population c The position vector of the parameter configuration; Represented as the first c The position vector of the parameter configuration; S Represented as a random number; P It represents the maximum value of the parameter configuration position vector in the initial population; H Represents the minimum value of the parameter configuration position vector in the initial population.

4. The predictive maintenance method for hydraulic equipment based on the Bayesian classification model according to claim 3 is characterized in that: The position update formula is: ； In the formula, represents the position vector of the parameter configuration in the next iteration; The position vector of the parameter configuration in the current iteration; Q Represented as parameter configuration; x It is expressed as the golden section number; Represented as a random number; Represented as a random number; W Expressed as an adjustment factor.

5. The predictive maintenance method for hydraulic equipment based on the Bayesian classification model according to claim 1 is characterized in that: The method of applying the load analysis algorithm to the optimization model to find the best solution that meets the constraint conditions and using it as the factor data affecting the load of the hydraulic equipment includes the following steps: S351, processing the identified factors affecting the load of the hydraulic equipment to ensure that the factors affecting the load of the hydraulic equipment meet the preset operating conditions, and obtaining the processed factor data; S352, initializing the parameters of the load analysis algorithm and setting the initial hydraulic equipment load optimization improvement technology; S353, evaluating the specific impact of each improvement technology on the load of the hydraulic equipment through the established objective function, and updating the status according to the evaluation results; S354, optimize, evaluate and adjust the improvement technology of factors affecting the load of hydraulic equipment, and continuously update the status according to the preset operating conditions; S355. If the load efficiency of the optimized hydraulic equipment reaches a preset threshold, the best improvement technology is determined as the factor data affecting the load of the hydraulic equipment. If the load efficiency of the optimized hydraulic equipment does not reach the preset threshold, the improvement technology is continued to be optimized and adjusted.

6. The predictive maintenance method for hydraulic equipment based on the Bayesian classification model according to claim 1 is characterized in that: The step of comparing the identified factor data affecting the load of the hydraulic equipment with the known hydraulic equipment failure factor data, determining the key factor affecting the failure of the hydraulic equipment, and debugging and maintaining the hydraulic equipment includes the following steps: S41, calculating the importance score of each factor affecting the load of the hydraulic equipment on the load of the hydraulic equipment according to the objective function, and selecting the factor that has the greatest impact on the failure of the hydraulic equipment as the key factor; S42, performing a global search, calculating the impact of key factors under different settings on the performance of the hydraulic equipment, and comparing with the current equipment configuration; S43. After each round of optimization, evaluate the new settings of all key factors and select the best equipment configuration as a new starting point for further optimization; S44, updating the environmental factor settings according to the operation feedback and efficiency data of the hydraulic equipment; S45, determining whether the optimization termination condition is met, if so, confirming the optimal setting, if not, returning to step S41 to continue optimization; S46. Use the optimized and confirmed settings as the final adjustment of the environmental and operating conditions of the hydraulic equipment to ensure that the equipment operates in the best condition.

7. The predictive maintenance method for hydraulic equipment based on the Bayesian classification model according to claim 6 is characterized in that: The formula for calculating the importance score of each factor affecting the load of the hydraulic equipment on the load of the hydraulic equipment according to the objective function is: ； In the formula, K It is expressed as the objective function; D Represented as an importance membership matrix; R It is expressed as a collection of factors that affect the load of hydraulic equipment; Expressed as factors affecting the load of hydraulic equipment z Key factors affecting hydraulic equipment failure h The importance of membership; e Expressed as the total number of factors affecting the load on the hydraulic equipment; l Expressed as the total number of key factors affecting hydraulic equipment failure; d It is expressed as fuzzy coefficient; Expressed as factors affecting the load of hydraulic equipment z Key factors affecting hydraulic equipment failure h The influence distance between them.

8. A hydraulic equipment predictive maintenance system based on a Bayesian classification model, used to implement the hydraulic equipment predictive maintenance method based on a Bayesian classification model as described in any one of claims 1 to 7, characterized in that: The predictive maintenance system for hydraulic equipment based on the Bayesian classification model includes: The data acquisition module is used to acquire the real-time operation status data of the hydraulic equipment and pre-process the real-time operation status data to obtain the operation status characteristic data; Establish a model module for bringing the operating status feature data into the Bayesian classification model for training to obtain a load prediction model, and use the load prediction model to predict the load state of the hydraulic equipment at a future moment to obtain the load data of the hydraulic equipment; A load analysis module is used to analyze the load data of the hydraulic equipment using a load analysis algorithm and identify the factor data that affects the load of the hydraulic equipment; The factor analysis module is used to compare the identified factor data affecting the load of hydraulic equipment with the known hydraulic equipment failure factor data, determine the key factors affecting the failure of hydraulic equipment, and perform debugging and maintenance on the hydraulic equipment; The data acquisition module is connected with the load analysis module by establishing a model module, and the load analysis module is connected with the factor analysis module.

Images