A noise reduction and optimization management method and system for compressed air systems

Abstract

This invention discloses a noise reduction optimization management method and system for compressed air systems, belonging to the field of noise reduction optimization technology. The method includes: integrating noise reduction-related data of the compressed air system, optimizing the noise reduction-related data of the compressed air system, constructing a noise reduction optimization management model, and noise reduction optimization management. This scheme obtains the dynamic reliability coefficient of the data on the features based on the heterogeneity index of the data relative to the features, calculates the weighted distance and constructs the nearest neighbor set to obtain the cluster deviation coefficient of the data, and identifies and deletes abnormal data. A modified harmonic series function is introduced to obtain the Euler-harmonic information entropy of the set. Based on the Euler gain and control factors, the optimal splitting feature and the optimal splitting feature value are selected to obtain a decision tree constructed based on the structure of the data subset. Then, based on the label definition of the data subset, the labels of the leaf nodes in the decision tree are redefined, which helps to perform real-time and dynamic noise reduction optimization management of the compressed air system, improving management efficiency and noise reduction effect.

Description

Technical Field

[0001] This invention belongs to the field of noise reduction and optimization technology, specifically referring to a noise reduction and optimization management method and system for compressed air systems. Background Technology

[0002] The noise reduction optimization management method utilizes artificial intelligence technology to collect and process multi-dimensional data on the noise reduction of compressed air systems, uncover potential patterns in historical data, optimize the noise reduction control scheme of compressed air systems, effectively reduce the noise level during the operation of compressed air systems, improve the operational stability and work efficiency of compressed air systems, create a quiet and comfortable working environment, reduce the impact of noise on the health and work status of surrounding personnel, and also help reduce the risk of equipment failure caused by noise and extend the service life of equipment.

[0003] However, existing noise reduction optimization management methods suffer from several drawbacks. During operation, due to sensor errors, environmental interference, and fluctuations in operating conditions, some noise reduction-related data collected from the compressed air system deviates from the normal range, resulting in low data quality. This reduces the accuracy of noise reduction optimization management and leads to poor noise reduction optimization effects. Furthermore, existing noise reduction optimization management methods are hampered by the highly complex and diverse nature of the collected noise reduction-related data. Traditional methods struggle to effectively adapt to the dynamic changes in this data, resulting in ineffective noise reduction strategies and hindering effective noise reduction optimization management of the compressed air system. Summary of the Invention

[0004] To address the above-mentioned issues and overcome the shortcomings of existing technologies, this invention provides a noise reduction optimization management method and system for compressed air systems. Addressing the problem that existing noise reduction optimization management methods suffer from low data quality due to sensor errors, environmental interference, and operating condition fluctuations during system operation, which can lead to some data deviating from the normal range and thus reducing the accuracy of noise reduction optimization management and resulting in poor noise reduction optimization effects, this solution utilizes the data's alienation index relative to features to obtain a dynamic reliability coefficient for the data on specific features. This allows for the identification of anomalies in individual data points on specific features. Weighted distance calculations based on the dynamic reliability coefficient improve distance accuracy and reduce false alarm rates. The construction of a nearest neighbor set yields the data's cluster deviation coefficient, identifying and deleting abnormal data, thus more accurately assessing data reliability, effectively eliminating outliers, and improving data quality. This makes subsequent data-driven noise reduction optimization management analysis more accurate. Furthermore, this solution addresses the issue that existing noise reduction optimization management methods suffer from low data quality due to sensor errors, environmental interference, and operating condition fluctuations during system operation, resulting in some data deviating from the normal range and thus poor noise reduction optimization effects. Noise reduction-related data is highly complex and diverse, making it difficult for traditional methods to effectively adapt to the dynamic changes in the data. This results in poor noise reduction strategies and hinders effective noise reduction optimization management of compressed air systems. This solution introduces a modified harmonic series function to obtain the Euler-harmonic information entropy of the set, which can more comprehensively and accurately measure the information content and uncertainty of the data subset. Based on Euler gain and control factors, the optimal splitting feature and optimal splitting feature value are selected to obtain a decision tree based on the structure of the data subset, which helps to build a more efficient and accurate decision tree. Furthermore, the labels of the leaf nodes in the decision tree are redefined based on the label definition of the data subset, making the output of the decision tree more in line with the requirements of noise reduction optimization management. Based on ensemble learning technology, the decision trees are combined to obtain a noise reduction optimization management model, which better copes with the high complexity and dynamic changes of compressed air system data. More targeted noise reduction strategies are formulated for different operating conditions, which helps to optimize noise reduction of the compressed air system in real time and dynamically, improving management efficiency and noise reduction effect.

[0005] The technical solution adopted by this invention is as follows: This invention provides a noise reduction and optimization management method for compressed air systems, which includes the following steps:

[0006] Step S1: Integrate noise reduction data related to the compressed air system;

[0007] Step S2: Optimize noise reduction data related to the compressed air system;

[0008] Step S3: Construct a noise reduction optimization management model;

[0009] Step S4: Noise Reduction Optimization Management.

[0010] Furthermore, in step S1, the integration of noise reduction-related data of the compressed air system involves collecting and processing historical noise reduction-related data of the compressed air system.

[0011] The collected historical noise reduction data of the compressed air system includes compressed air system operation data, compressed air system equipment data, noise data, environmental data, and noise reduction adjustment data.

[0012] The process of processing historical compressed air system noise reduction data involves data cleaning, data normalization, data encoding, and dataset construction to obtain a compressed air system noise reduction dataset.

[0013] Furthermore, in step S2, the optimization of the noise reduction related data of the compressed air system specifically includes the following steps:

[0014] Step S21: Calculate the alienation index; calculate the Euclidean distance between every two data points in the noise reduction dataset A of the compressed air system on the same feature, and then normalize the data based on the maximum and minimum Euclidean distances to obtain the alienation index of the data relative to the feature; the formula used is as follows:

[0015] ;

[0016] In the formula, and These are the maximum and minimum Euclidean distances between data v in A and the other data on feature m, respectively. It is the heterogeneity index of data v relative to feature m, where v and w are data indices in A, and m is the feature index in A. and These are the projections of data v and data w onto feature m, respectively. yes and The Euclidean distance between them, where H is the number of data in A;

[0017] Step S22: Calculate the dynamic reliability coefficient; calculate the average alienation index of all data in A on the same feature, and apply the average alienation index to... The standardization process is performed to obtain the standardized alienation index, and then the dynamic confidence coefficient of the feature is calculated using the inverse Sigmoid function.

[0018] Step S23: Calculate the nearest neighbor data; based on the dynamic confidence coefficient, calculate the weighted distance between data v in A and each of the other data, sort the weighted distances in ascending order, and select the top... The data corresponding to the smaller weighted distance is taken as the nearest neighbor data of data v, thus obtaining the nearest neighbor set of each data in A; where, It is rounded down.

[0019] Step S24: Calculate the cluster deviation coefficient; calculate the cluster deviation coefficient of the data based on the weighted distance and the nearest neighbor set;

[0020] Step S25: Optimization; Pre-set deviation threshold, delete data with a cluster deviation coefficient greater than the deviation threshold as abnormal data from A, and obtain the optimized management dataset. Select 70% of the data from the optimized management dataset to construct the training dataset, and use the remaining 30% of the data as the test dataset.

[0021] Furthermore, in step S3, the construction of the noise reduction optimization management model specifically includes the following steps:

[0022] Step S31: Training dataset partitioning; randomly partition the training dataset into two non-overlapping subsets Z1 and Z2, Z1 and Z2 being the data subset for structure construction and the data subset for label definition, respectively. Record the ratio of the number of data in Z1 and Z2, and set the values ​​of the parameter group [β1, β2, p]; where β1 and β2 are the first control factor and the second control factor, respectively, and p is the threshold.

[0023] Step S32: Calculate the Euler-harmonic information entropy; design a modified harmonic series function based on Euler's constant and the harmonic series, and introduce the modified harmonic series function to obtain the Euler-harmonic information entropy of the set; the formula used is as follows:

[0024] ;

[0025] in, It is the Euler-harmonic information entropy of the data subset Z1, where |Z1| is the number of data in Z1, Q is the number of data labels, q is the data label index, and h is the data label index. q It is the number of data belonging to data label q in Z1. It is a modified harmonic series function. It is constructed based on harmonic series and corrected by incorporating Euler's constant;

[0026] Step S33: Construct a decision tree; construct a binary decision tree based on the data subset Z1, and split each node of the decision tree according to the node splitting rules until the number of leaf nodes in the decision tree equals the threshold p, thus completing the construction of the decision tree; the node splitting rules include the following steps:

[0027] Step S331: Calculate the Euler gain; subtract the weighted Euler-harmonic information entropy of the data sets of the left and right child nodes from the Euler-harmonic information entropy of the data set at the node to obtain the Euler gain;

[0028] Step S332: Select the best splitting feature; construct the maximum Euler gain set when splitting with each feature at the node, and normalize the maximum Euler gain set to obtain the normalized maximum Euler gain set. Calculate the selection probability of each feature in the normalized maximum Euler gain set based on the first control factor β1 to obtain the feature selection probability set. Randomly select a feature as the best splitting feature according to the feature selection probability set.

[0029] Step S333: Select the optimal splitting feature value; construct the Euler gain set when splitting at the node using all possible feature values ​​of the optimal splitting feature, and normalize the Euler gain set to obtain the normalized Euler gain set. Calculate the selection probability of each feature value in the normalized Euler gain set based on the second control factor β2 to obtain the feature value selection probability set. Then, according to the feature value selection probability set, randomly select a feature value from all possible feature values ​​of the optimal splitting feature as the optimal splitting feature value.

[0030] Step S334: Split; Using the selected optimal splitting feature value, split the node to obtain its left and right child nodes;

[0031] Step S34: Redefine labels; input the data in the label definition data subset Z2 into the constructed decision tree, allocate the data to the corresponding leaf nodes according to the node splitting rules, obtain the data set corresponding to each leaf node on Z2, calculate the prediction probability of each data label in the data set, obtain the prediction probability set, and then select the data label with the highest prediction probability from all data labels as the new data label of the leaf node according to the prediction probability set.

[0032] Step S35: Decision tree combination; Repeat steps S31-S34 N times to construct a total of N decision trees. Based on ensemble learning technology, combine the N decision trees to obtain a noise reduction and optimization management model, and output the predicted data labels.

[0033] Further, in step S4, the noise reduction optimization management involves collecting and processing real-time noise reduction related data of the compressed air system, inputting the processed real-time noise reduction related data of the compressed air system into the noise reduction optimization management model for analysis, obtaining the corresponding noise reduction adjustment data according to the output data labels, generating a noise reduction adjustment strategy, and performing noise reduction optimization management on the compressed air system according to the noise reduction adjustment strategy.

[0034] The data collected in real time related to noise reduction of the compressed air system includes compressed air system operation data, compressed air system equipment data, noise data, and environmental data.

[0035] The process of processing real-time compressed air system noise reduction related data involves data cleaning, data normalization, and data encoding.

[0036] This invention provides a noise reduction optimization management system for compressed air systems, comprising a compressed air system noise reduction related data integration module, a compressed air system noise reduction related data optimization module, a noise reduction optimization management model construction module, and a noise reduction optimization management module;

[0037] The compressed air system noise reduction related data integration module collects and processes historical compressed air system noise reduction related data, and sends the data to the compressed air system noise reduction related data optimization module;

[0038] The compressed air system noise reduction related data optimization module obtains the dynamic reliability coefficient of the data on the feature based on the alienation index of the data relative to the feature, calculates the weighted distance and constructs the nearest neighbor set to obtain the cluster deviation coefficient of the data, identifies and deletes abnormal data, and sends the data to the noise reduction optimization management model construction module.

[0039] The noise reduction optimization management model construction module introduces a modified harmonic series function to obtain the Euler-harmonic information entropy of the set. Based on the Euler gain and control factor, the optimal splitting feature and the optimal splitting feature value are selected to obtain a decision tree constructed based on the structure-based data subset. Then, based on the label definition data subset, the labels of the leaf nodes in the decision tree are redefined. Based on the ensemble learning technique, the decision tree is combined to obtain the noise reduction optimization management model, and the data is sent to the noise reduction optimization management module.

[0040] The noise reduction optimization management module obtains the corresponding noise reduction adjustment data based on the data tags output by the noise reduction optimization management model, generates a noise reduction adjustment strategy, and performs noise reduction optimization management on the compressed air system according to the noise reduction adjustment strategy.

[0041] The beneficial effects achieved by the present invention using the above solution are as follows:

[0042] (1) In response to the problem that some data in the noise reduction-related data of the compressed air system deviates from the normal range due to sensor errors, environmental interference and operating condition fluctuations during the operation of the compressed air system, resulting in low data quality, reduced accuracy of noise reduction optimization management and poor noise reduction optimization effect of the compressed air system, this solution obtains the dynamic reliability coefficient of the data on the feature based on the alienation index of the data relative to the feature, which can identify the anomaly of a single data point on a specific feature; calculates the weighted distance based on the dynamic reliability coefficient, making the distance calculation more accurate and reducing the false alarm rate; constructs the nearest neighbor set to obtain the cluster deviation coefficient of the data, identifies and deletes abnormal data, more accurately identifies the reliability of the data, effectively removes outliers, improves data quality, and makes the subsequent data-based noise reduction optimization management analysis more accurate, ultimately achieving more effective noise control.

[0043] (2) To address the problem that existing noise reduction optimization management methods suffer from highly complex and diverse noise reduction-related data of compressed air systems, making it difficult for traditional methods to effectively adapt to the dynamic changes in the data and resulting in poor noise reduction strategy performance and ineffective noise reduction optimization management of compressed air systems, this scheme randomly divides the training dataset into a structured data subset and a label-defined data subset. This division allows for more systematic mining of information from the data. The introduction of a modified harmonic series function to obtain the Euler-harmonic information entropy of the set provides a more comprehensive and accurate measurement of the information content and uncertainty of the data subset, taking into account the dynamic changes in the data and overcoming the problem of inaccurate information measurement in traditional methods when dealing with complex dynamic data. Based on the Euler-harmonic series function... By selecting the optimal splitting feature and optimal splitting feature value based on the benefit and control factors, a decision tree constructed based on the structured data subset is obtained, which helps to build a more efficient and accurate decision tree. Then, based on the label definition data subset, the labels of the leaf nodes in the decision tree are redefined, so that the output of the decision tree is more in line with the requirements of noise reduction optimization management, ensuring that the decision tree can accurately predict appropriate noise reduction strategies based on the input data. Based on ensemble learning technology, the decision tree is combined to obtain a noise reduction optimization management model, which improves the generalization ability and stability, better copes with the high complexity and dynamic change characteristics of compressed air system data, and formulates more targeted noise reduction strategies for different operating conditions. This helps to perform real-time and dynamic noise reduction optimization management of the compressed air system, improving management efficiency and noise reduction effect. Attached Figure Description

[0044] Figure 1 A flowchart illustrating a noise reduction and optimization management method for a compressed air system provided by the present invention;

[0045] Figure 2 A schematic diagram of a noise reduction and optimization management system for a compressed air system provided by the present invention;

[0046] Figure 3 This is a flowchart illustrating step S2;

[0047] Figure 4 This is a flowchart illustrating step S3.

[0048] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. Detailed Implementation

[0049] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0050] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0051] Example 1, see Figure 1 The present invention provides a noise reduction and optimization management method for compressed air systems, the method comprising the following steps:

[0052] Step S1: Integrate noise reduction data related to the compressed air system; collect and process historical noise reduction data related to the compressed air system;

[0053] Step S2: Optimize noise reduction related data of compressed air system; Based on the alienation index of data relative to features, obtain the dynamic credibility coefficient of data on features, calculate the weighted distance and construct the nearest neighbor set to obtain the cluster deviation coefficient of data, and identify and delete abnormal data;

[0054] Step S3: Construct a noise reduction optimization management model; introduce a modified harmonic series function to obtain the Euler-harmonic information entropy of the set, select the best splitting feature and the best splitting feature value according to the Euler gain and control factor, and obtain a decision tree based on the data subset constructed by the structure. Then, redefine the labels of the leaf nodes in the decision tree based on the label definition data subset, and combine the decision trees to obtain the noise reduction optimization management model.

[0055] Step S4: Noise Reduction Optimization Management; Based on the data labels output by the noise reduction optimization management model, obtain the corresponding noise reduction adjustment data, generate a noise reduction adjustment strategy, and perform noise reduction optimization management on the compressed air system according to the noise reduction adjustment strategy.

[0056] Example 2, see Figure 1 This embodiment is based on the above embodiment. In step S1, the integration of noise reduction related data of the compressed air system is to collect and process historical noise reduction related data of the compressed air system.

[0057] The collected historical noise reduction data of the compressed air system includes compressed air system operation data, compressed air system equipment data, noise data, environmental data, and noise reduction adjustment data.

[0058] The compressed air system operating data includes fan speed, air pressure, air volume, fan vibration intensity, and operating time;

[0059] The compressed air system equipment data includes the fan model, number of blades, maintenance record data, and failure rate;

[0060] The noise data includes noise intensity and noise frequency;

[0061] The environmental data includes temperature, humidity, atmospheric pressure, wind speed, and wind direction;

[0062] The process of processing historical compressed air system noise reduction data involves data cleaning, data normalization, data encoding, and dataset construction to obtain a compressed air system noise reduction dataset.

[0063] The data cleaning process involves removing erroneous, missing, and outlier values ​​from the data.

[0064] The data normalization method uses the max-min scaling method to unify numerical data to the same range.

[0065] The data encoding uses One-Hot encoding to convert categorical data into numerical data.

[0066] The aforementioned dataset construction involves using noise reduction adjustment data as data labels, and then constructing a noise reduction dataset for the compressed air system based on historical compressed air system noise reduction-related data after data cleaning, data normalization, and data encoding.

[0067] Example 3, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S2, the optimization of noise reduction related data of the compressed air system specifically includes the following:

[0068] Step S21: Calculate the alienation index; calculate the Euclidean distance between every two data points in the noise reduction dataset A of the compressed air system on the same feature, and then normalize the data based on the maximum and minimum Euclidean distances to obtain the alienation index of the data relative to the feature; by calculating the Euclidean distance between data points on the same feature, outliers that significantly deviate from other data on specific features are identified, and normalization is used to eliminate the dimensional differences between different features, making subsequent analysis more balanced; the formula used is as follows:

[0069] ;

[0070] ;

[0071] ;

[0072] In the formula, and These are the maximum and minimum Euclidean distances between data v in A and the other data on feature m, respectively. It is the heterogeneity index of data v relative to feature m, where v and w are data indices in A, and m is the feature index in A. and These are the projections of data v and data w onto feature m, respectively. yes and The Euclidean distance between them, where H is the number of data in A;

[0073] Step S22: Calculate the dynamic reliability coefficient; calculate the average alienation index of all data in A on the same feature, and apply the average alienation index to... Standardization is performed to obtain the standardized alienation index, and then the dynamic confidence coefficient of the feature is calculated using the inverse sigmoid function. The inverse sigmoid function assigns a low confidence coefficient to features with a high alienation index and a high confidence coefficient to features with a low alienation index. The formula used is as follows:

[0074] ;

[0075] ;

[0076] ;

[0077] in, It is the alienation index of data v relative to feature m. It is the average alienation index of all data in A on feature m. It is the standardized alienation index of data v relative to feature m. is the standard deviation of the heterogeneity index of all data in A on feature m, v is the data index in A, and m is the feature index in A. It is the dynamic confidence coefficient of data i on feature m;

[0078] Step S23: Calculate the nearest neighbor data; based on the dynamic confidence coefficient, calculate the weighted distance between data v in A and each of the other data, sort the weighted distances in ascending order, and select the top... The data corresponding to the smaller weighted distance is used as the nearest neighbor data of data v, thus obtaining the nearest neighbor set of each data in A; the neighbor data most similar to the target data is found by weighted distance, providing a local reference benchmark for subsequent anomaly detection; the formula used is as follows:

[0079] ;

[0080] In the formula, M is the weighted distance between data v and data w in A, and M is the number of features in A. It rounds down;

[0081] Step S24: Calculate the cluster deviation coefficient; calculate the cluster deviation coefficient of the data based on the weighted distance and the nearest neighbor set; identify isolated anomalies or group anomalies by comparing the sum of the weighted distances between the data and its neighbors, and introduce μ. o To eliminate the scaling effect and make the cluster deviation coefficients comparable under different conditions, the following formula was used:

[0082] ;

[0083] In the formula, D v C is the cluster skewness coefficient of data v. v C is the weighted sum of distances between data i and all data in its nearest neighbor set. o It is the weighted sum of the distances between data i's nearest neighbor data o and all data in its own nearest neighbor set. μ is the weighted distance between data v in A and its nearest neighbor data o. o It is the weighted average distance between the nearest neighbor data o and all data in its own nearest neighbor set, where o is the nearest neighbor data index;

[0084] Step S25: Optimization; Pre-set deviation threshold, delete data with a cluster deviation coefficient greater than the deviation threshold as abnormal data from A to obtain the optimized management dataset, select 70% of the data from the optimized management dataset to construct the training dataset, and use the remaining 30% of the data as the test dataset; improve the quality of the dataset by removing abnormal data through deviation threshold.

[0085] By performing the above operations, this solution addresses the problem in existing noise reduction optimization management methods where some data in the noise reduction-related data of the compressed air system deviates from the normal range due to sensor errors, environmental interference, and operating condition fluctuations during operation. This results in low data quality, reduced accuracy of noise reduction optimization management, and poor noise reduction optimization effects. This solution obtains the dynamic reliability coefficient of the data on the feature based on the alienation index of the data relative to the feature, enabling the identification of anomalies in specific features for individual data points. Weighted distance is calculated based on the dynamic reliability coefficient, making distance calculation more accurate and reducing the false alarm rate. A nearest neighbor set is constructed to obtain the cluster deviation coefficient of the data, identifying and deleting abnormal data, more accurately identifying data reliability, effectively eliminating outliers, and improving data quality. This makes subsequent data-based noise reduction optimization management analysis more accurate, ultimately achieving more effective noise control.

[0086] Example 4, see Figure 1 and Figure 4 This embodiment is based on the above embodiment. In step S3, the construction of the noise reduction optimization management model specifically includes the following steps:

[0087] Step S31: Training dataset partitioning; randomly partition the training dataset into two non-overlapping subsets Z1 and Z2, where Z1 is the data subset for structure construction and Z2 is the data subset for label definition, respectively. Record the ratio of the number of data in Z1 to Z2. The parameter set [β1, β2, p] is set, where β1 and β2 are the first and second control factors, respectively, p is the threshold, and |Z1| and |Z2| are the number of data points in Z1 and Z2, respectively. Z1 is used to construct the structure of the decision tree, and Z2 is used to redefine the labels of the leaf nodes in the decision tree. Z2 is independent of Z1 and can correct the label bias of the decision tree, making the model more adaptable to the real situation. By controlling the model structure through the ratio of data volume and the parameter set, a single data distribution is avoided from dominating the construction of the decision tree.

[0088] Step S32: Calculate the Euler-harmonic information entropy; design a modified harmonic series function based on the Euler constant and harmonic series, and introduce the modified harmonic series function to obtain the Euler-harmonic information entropy of the set; making it more suitable for the statistical characteristics of noise reduction related data of the compressed air system; the formula used is as follows:

[0089] ;

[0090] in, It is the Euler-harmonic information entropy of the data subset Z1, where |Z1| is the number of data in Z1, Q is the number of data labels, q is the data label index, and h is the data label index. q It is the number of data belonging to data label q in Z1. It is a modified harmonic series function. It is constructed based on harmonic series and corrected by incorporating Euler's constant. γ is Euler's constant;

[0091] Step S33: Construct a decision tree; construct a binary decision tree based on the data subset Z1, and split each node of the decision tree according to the node splitting rules until the number of leaf nodes in the decision tree equals the threshold p, thus completing the construction of the decision tree; the node splitting rules include the following steps:

[0092] Step S331: Calculate the Euler gain; subtract the weighted Euler-harmonic entropy of the data sets of the left and right child nodes from the Euler-harmonic entropy of the data set at the node to obtain the Euler gain; the formula used is as follows:

[0093] ;

[0094] In the formula, Z1 is the Euler gain corresponding to splitting at node a using the feature value j of feature i. a It is the data set at node a of the decision tree built based on Z1. Let j be the feature value of feature i, where a, i, and j are the node index of the decision tree, the feature index in Z1, and the feature value index in Z1, respectively. and They are used at node a respectively The data sets of its left and right child nodes after splitting. , and These are sets Z1 a ,gather and set Euler-harmonic information entropy, , and These are sets Z1 a ,gather and set The amount of data in and They are using b at node a respectively ij The weighted Euler-harmonic information entropy of the data sets of the left and right child nodes after splitting;

[0095] Step S332: Select the optimal splitting feature; construct the set of maximum Euler gains when splitting using each feature at the node, and normalize the set of maximum Euler gains to obtain the normalized set of maximum Euler gains. Calculate the selection probability of each feature in the normalized set of maximum Euler gains based on the first control factor β1 to obtain the feature selection probability set. Randomly select a feature as the optimal splitting feature based on the feature selection probability set; improve generalization ability so that the model can adapt to noise reduction data related to compressed air systems under different operating conditions; the formula used is as follows:

[0096] ;

[0097] ;

[0098] ;

[0099] ;

[0100] Among them, L a and These are the set of maximum Euler gains and the set of normalized maximum Euler gains at node a, respectively, where I is the total number of features. , and These are the maximum Euler gains generated when splitting at node a using all possible feature values ​​of the 1st, 1st, and 1st features, respectively. a It is the set of feature selection probabilities at node a. It is the softmax function; At node a, for all possible feature values ​​j of feature i, calculate the corresponding Euler gain. Then take the maximum value; At node a, the maximum Euler gain for all features i. Take the minimum value; At node a, the maximum Euler gain for all features i. Take the maximum value;

[0101] Step S333: Select the optimal splitting eigenvalue; construct a set of Euler gains for splitting at nodes using all possible eigenvalues ​​of the optimal splitting feature, and normalize the Euler gain set to obtain a normalized Euler gain set. Calculate the selection probability of each eigenvalue in the normalized Euler gain set based on the second control factor β2 to obtain a eigenvalue selection probability set. Then, based on the eigenvalue selection probability set, randomly select one eigenvalue from all possible eigenvalues ​​of the optimal splitting feature as the optimal splitting eigenvalue; automatically adapt to the noise characteristics of different compressed air systems; the formula used is as follows:

[0102] ;

[0103] ;

[0104] ;

[0105] In the formula, and These are the Euler gain set and the normalized Euler gain set when splitting node a using the optimal splitting feature ibest, respectively. ibest is the index of the optimal splitting feature, and J is the total number of feature values ​​for the optimal splitting feature. , and These are the 1st, jth, and Jth feature values ​​of the optimal splitting feature ibest, respectively. , and These represent the Euler gains at node a when splitting using the 1st, jth, and Jth feature values ​​of the optimal splitting feature ibest. It is the set of feature value selection probabilities when splitting at node a using the best splitting feature ibest. At node a, for all possible feature values ​​j of the selected best splitting feature ibest, calculate the corresponding Euler gain, and then take the minimum value among them; At node a, for all possible feature values ​​j of the selected best splitting feature ibest, calculate the corresponding Euler gain, and then take the maximum value among them;

[0106] Step S334: Split; Using the selected optimal splitting feature value, split the node to obtain its left and right child nodes;

[0107] Step S34: Redefine labels; input the data from the label definition data subset Z2 into the constructed decision tree, and allocate the data to the corresponding leaf nodes according to the node splitting rules to obtain the data set corresponding to each leaf node on Z2. Calculate the prediction probability of each data label in the data set to obtain the prediction probability set. Then, based on the prediction probability set, select the data label with the highest prediction probability from all data labels as the new data label for the leaf node. The leaf node labels of the decision tree may be incorrect due to data bias in Z1. Z2 corrects the label bias, ensuring that the model can still predict correctly under rare conditions, avoiding label oscillations caused by training data noise, and improving model stability. The formula used is as follows:

[0108] ;

[0109] ;

[0110] Among them, G c Z2 is the set of predicted probabilities for leaf node c. c It is the data set corresponding to leaf node c on Z2, |Z2 c |It's Z2 c The number of data points in the decision tree, where c is the index of a leaf node. , and The leaf node c is in Z2. c The predicted probabilities of the 1st, qth, and Qth data labels, where Q is the number of data labels, q is the data label index, and x is the predicted probability of the 1st, qth, and Qth data labels. d It's Z2 c The d-th data point in Z2, where d is the value of Z2. c Data index in It is a decision tree for data x d Predicted data labels; It is an indicator function; when hour, =1, otherwise =0;

[0111] Step S35: Decision Tree Combination; Repeat steps S31-S34 N times to construct a total of N decision trees. Based on ensemble learning technology, combine the N decision trees to obtain the noise reduction optimization management model and output the predicted data labels. A single decision tree may be affected by randomness, leading to unstable noise reduction strategies. By using multiple decision trees to make decisions together, the model can be applied to different noise reduction related data of compressed air systems. The formula used is as follows:

[0112] ;

[0113] Where Y is the predicted data label output by the noise reduction optimization management model. It is the predicted data label of the nth decision tree for data x; It is an indicator function; when hour, =1, otherwise It is 0.

[0114] By performing the above operations, this solution addresses the problem in existing noise reduction optimization management methods where the collected noise reduction-related data for compressed air systems is highly complex and diverse. Traditional methods struggle to effectively adapt to the dynamic changes in this data, leading to poor noise reduction strategies and hindering effective noise reduction optimization management of compressed air systems. This solution randomly divides the training dataset into a structured data subset and a label-defined data subset. This division allows for more systematic information mining within the data. Furthermore, by introducing a modified harmonic series function to obtain the Euler-harmonic information entropy of the set, it can more comprehensively and accurately measure the information content and uncertainty of the data subset, taking into account the dynamic changes in the data and overcoming the problem of inaccurate information measurement in traditional methods when dealing with complex dynamic data. Based on... Euler gain and control factors are used to select the optimal splitting feature and optimal splitting feature value, resulting in a decision tree constructed based on a data subset with a structured structure. This helps to build a more efficient and accurate decision tree. Then, the labels of the leaf nodes in the decision tree are redefined based on the labeled data subset, making the output of the decision tree more consistent with the requirements of noise reduction optimization management. This ensures that the decision tree can accurately predict appropriate noise reduction strategies based on the input data. Based on ensemble learning technology, the decision trees are combined to obtain a noise reduction optimization management model, improving generalization ability and stability. This model better copes with the high complexity and dynamic changes of compressed air system data, allowing for more targeted noise reduction strategies to be developed for different operating conditions. This facilitates real-time, dynamic noise reduction optimization management of the compressed air system, improving management efficiency and noise reduction effectiveness.

[0115] Example 5, see Figure 1 This embodiment is based on the above embodiment. In step S4, the noise reduction optimization management involves collecting and processing real-time noise reduction related data of the compressed air system, inputting the processed real-time noise reduction related data of the compressed air system into the noise reduction optimization management model for analysis, obtaining the corresponding noise reduction adjustment data according to the output data label, generating a noise reduction adjustment strategy, and performing noise reduction optimization management on the compressed air system according to the noise reduction adjustment strategy.

[0116] The data collected in real time related to noise reduction of the compressed air system includes compressed air system operation data, compressed air system equipment data, noise data, and environmental data.

[0117] The process of processing real-time compressed air system noise reduction related data involves data cleaning, data normalization, and data encoding.

[0118] Example 6, see Figure 2 Based on the above embodiments, this embodiment provides a noise reduction optimization management system for compressed air systems, including a compressed air system noise reduction related data integration module, a compressed air system noise reduction related data optimization module, a noise reduction optimization management model construction module, and a noise reduction optimization management module.

[0119] The compressed air system noise reduction related data integration module collects and processes historical compressed air system noise reduction related data, and sends the data to the compressed air system noise reduction related data optimization module;

[0120] The compressed air system noise reduction related data optimization module obtains the dynamic reliability coefficient of the data on the feature based on the alienation index of the data relative to the feature, calculates the weighted distance and constructs the nearest neighbor set to obtain the cluster deviation coefficient of the data, identifies and deletes abnormal data, and sends the data to the noise reduction optimization management model construction module.

[0121] The noise reduction optimization management model construction module introduces a modified harmonic series function to obtain the Euler-harmonic information entropy of the set. Based on the Euler gain and control factor, the optimal splitting feature and the optimal splitting feature value are selected to obtain a decision tree constructed based on the structure-based data subset. Then, based on the label definition data subset, the labels of the leaf nodes in the decision tree are redefined. Based on the ensemble learning technique, the decision tree is combined to obtain the noise reduction optimization management model, and the data is sent to the noise reduction optimization management module.

[0122] The noise reduction optimization management module obtains the corresponding noise reduction adjustment data based on the data tags output by the noise reduction optimization management model, generates a noise reduction adjustment strategy, and performs noise reduction optimization management on the compressed air system according to the noise reduction adjustment strategy.

[0123] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0124] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

[0125] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A method for noise reduction optimization management for a compressed air system, the method comprising: The method comprises the following steps: ​ Step S1: noise reduction related data of the compressed air system is integrated; historical noise reduction related data of the compressed air system is collected and processed; Step S2: noise reduction related data of the compressed air system is optimized; a dynamic confidence coefficient of the data on a feature is obtained according to a dissimilation index of the data relative to the feature, a weighted distance is calculated, a nearest neighbor set is constructed, a cluster deviation coefficient of the data is obtained, and abnormal data is identified and deleted; Step S3: a noise reduction optimization management model is constructed; a Euler-harmonic information entropy of the set is obtained by introducing a modified harmonic series function, a best split feature and a best split feature value are selected according to a Euler gain and a control factor, a decision tree constructed based on a structure is obtained, a label of each leaf node in the decision tree is redefined based on a label definition data subset, the decision tree is combined, and a noise reduction optimization management model is obtained; Step S4: noise reduction optimization management is performed; corresponding noise reduction adjustment data is obtained according to a data label output by the noise reduction optimization management model, a noise reduction adjustment strategy is generated, and the compressed air system is managed in a noise reduction optimization manner according to the noise reduction adjustment strategy; In step S2, the noise reduction related data of the compressed air system is optimized, specifically comprising the following steps: Step S21: a dissimilation index is calculated; the formula used is as follows: ； wherein, and are the maximum and minimum Euclidean distances of data v in the noise reduction dataset A of the compressed air system and the rest of the data on feature m, respectively, is the dissimilarity index of data v with respect to feature m, v and w are data indices in A, and m is a feature index in A, and are the projections of data v and data w on feature m, respectively, is and the Euclidean distance between and H is the number of data in A; Step S22: Calculate the dynamic trust coefficient; calculate the average alienation index of all data in A on the same feature, and normalize the average alienation index to obtain the normalized alienation index, and then use the inverse Sigmoid function to calculate the dynamic trust coefficient of the feature; wherein A is a noise reduction data set of the compressed air system, is the alienation index of data v relative to feature m, v is the data index in A, and m is the feature index in A; is the alienation index of data v relative to feature m, v is the data index in A, and m is the feature index in A; Step S23: calculating the nearest neighbor data; calculating the weighted distance between the data v in A and each of the remaining data based on the dynamic trust coefficient, and sorting the weighted distances in ascending order, selecting the data corresponding to the first smaller weighted distances as the nearest neighbor data of the data v, thereby obtaining the nearest neighbor set of each data in A; wherein H is the number of data in A, is the floor function. Step S24: a cluster deviation coefficient is calculated; the cluster deviation coefficient of the data is calculated according to the weighted distance and the nearest neighbor set; Step S25: optimization; a deviation threshold is set in advance, data with a cluster deviation coefficient greater than the deviation threshold is deleted from A as abnormal data, an optimized management data set is obtained, 70% of the data in the optimized management data set is selected to construct a training data set, and the remaining 30% of the data is used as a test data set.

2. The method for noise reduction optimization management for a compressed air system of claim 1, wherein: In step S3, the noise reduction optimization management model is constructed, specifically comprising the following steps: Step S31: training data set division; the training data set is randomly divided into two non-overlapping subsets Z1 and Z2, Z1 and Z2 are a structure construction data subset and a label definition data subset respectively, the data quantity ratio of Z1 to Z2 is recorded, and the value of a parameter group [β1, β2, p] is set; wherein β1 and β2 are a first control factor and a second control factor respectively, and p is a threshold value; Step S32: Euler-harmonic information entropy calculation; a modified harmonic series function is designed based on Euler's constant and a harmonic series, and the Euler-harmonic information entropy of the set is obtained by introducing the modified harmonic series function; Step S33: decision tree construction; a decision tree with a binary structure is constructed based on the structure construction data subset Z1, each node of the decision tree is split based on a node splitting rule, and the node splitting is stopped when the number of leaf nodes in the decision tree is equal to the threshold value p, and the construction of the decision tree is completed; Step S34: redefinition of the label; the data in the label definition data subset Z2 is input into the constructed decision tree, the data is allocated to the corresponding leaf node according to the node splitting rule, the corresponding data set of each leaf node on Z2 is obtained, the prediction probability of each data label in the data set is calculated, the prediction probability set is obtained, and then the data label with the maximum prediction probability is selected from all data labels as the new data label of the leaf node according to the prediction probability set; Step S35: decision tree combination; repeat steps S31-S34 N times to construct N decision trees, combine the N decision trees based on ensemble learning technology, obtain a denoising optimization management model, and output a predicted data label.

3. The method for noise reduction optimization management for a compressed air system of claim 2, wherein: In step S32, the calculation of the Euler-Harmonic information entropy is based on the Euler constant and the modified harmonic series function designed by introducing the modified harmonic series function to obtain the Euler-Harmonic information entropy of the set; the formula used is as follows: ； wherein, is the Euler-Harmonic information entropy of the structure building data subset Z1, |Z1| is the number of data in Z1, Q is the number of data labels, q is the data label index, h q is the number of data in Z1 belonging to data label q, is the modified harmonic series function, is constructed based on the harmonic series and modified by the Euler constant.

4. The method for noise reduction optimization management for a compressed air system of claim 2, wherein: In step S33, the node splitting rule specifically includes the following steps: Step S331: Calculate the Euler gain; subtract the weighted Euler-Harmonic information entropy of the data set of the left and right child nodes from the Euler-Harmonic information entropy of the data set at the node, respectively, to obtain the Euler gain; Step S332: Select the best split feature; construct a set of maximum Euler gains when splitting using each feature at the node, and normalize the set of maximum Euler gains to obtain a set of normalized maximum Euler gains, calculate the selection probability of each feature in the set of normalized maximum Euler gains based on the first control factor β1 to obtain a set of feature selection probabilities, and randomly select a feature as the best split feature according to the set of feature selection probabilities; Step S333: Select the best split feature value; construct a set of Euler gains when splitting using all possible feature values of the best split feature at the node, and normalize the set of Euler gains to obtain a set of normalized Euler gains, calculate the selection probability of each feature value in the set of normalized Euler gains based on the second control factor β2 to obtain a set of feature value selection probabilities, and then randomly select a feature value from all possible feature values of the best split feature as the best split feature value according to the set of feature value selection probabilities; Step S334: Split; split the node using the selected best split feature value to obtain its left and right child nodes.

5. The method for noise reduction optimization management for a compressed air system of claim 1, wherein: In step S1, the integration of the pressure wind system denoising related data is to collect and process historical pressure wind system denoising related data; The collection of historical pressure wind system denoising related data is to collect pressure wind system operation data, pressure wind system equipment data, noise data, environmental data, and denoising adjustment data; The processing of historical pressure wind system denoising related data is to process the historical pressure wind system denoising related data through data cleaning, data normalization, data encoding, and data set construction to obtain a pressure wind system denoising data set.

6. The method for noise reduction optimization management for a compressed air system of claim 1, wherein: In step S4, the denoising optimization management is to collect and process real-time pressure wind system denoising related data, input the processed real-time pressure wind system denoising related data into the denoising optimization management model for analysis, obtain corresponding denoising adjustment data according to the output data label, generate a denoising adjustment strategy, and perform denoising optimization management on the pressure wind system according to the denoising adjustment strategy; The collection of real-time pressure wind system denoising related data is to collect pressure wind system operation data, pressure wind system equipment data, noise data, and environmental data; The processing of real-time pressure wind system denoising related data is to process the real-time pressure wind system denoising related data through data cleaning, data normalization, and data encoding.

7. A noise reduction optimization management system for a compressed air system for implementing a method of noise reduction optimization management for a compressed air system according to any one of claims 1 to 6, characterized in that: The pressure wind system noise reduction related data integration module, the pressure wind system noise reduction related data optimization module, the noise reduction optimization management model construction module and the noise reduction optimization management module are included. The pressure wind system noise reduction related data integration module collects and processes historical pressure wind system noise reduction related data and sends the data to the pressure wind system noise reduction related data optimization module. The pressure wind system noise reduction related data optimization module obtains a dynamic confidence coefficient of data on a feature according to a dissimilation index of the data relative to the feature, calculates a weighted distance and constructs a nearest neighbor set, obtains a cluster deviation coefficient of the data, identifies and deletes abnormal data, and sends the data to the noise reduction optimization management model construction module. The noise reduction optimization management model construction module introduces a modified harmonic series function to obtain an Euler-harmonic information entropy of the set, selects a best split feature and a best split feature value according to an Euler gain and a control factor, obtains a decision tree constructed based on a structure constructed data subset, redefines a label of a leaf node in the decision tree based on a label defined data subset, combines the decision trees based on an integrated learning technology, obtains a noise reduction optimization management model, and sends the data to the noise reduction optimization management module. The noise reduction optimization management module obtains corresponding noise reduction adjustment data according to a data label output by the noise reduction optimization management model, generates a noise reduction adjustment strategy, and performs noise reduction optimization management on the pressure wind system according to the noise reduction adjustment strategy.

Images