Reactor cooler performance evaluation and predictive maintenance method and device

Abstract

The invention provides a reactor cooler performance evaluation and predictive maintenance method and device, and relates to the technical field of performance evaluation, and the method comprises the steps: collecting the historical data of a reactor cooler in real time; obtaining an input variable feature set; inputting a prediction model, and cooperatively predicting cooling efficiency and future energy consumption data in combination with real-time monitoring data to obtain a performance degradation curve of the reactor cooler; calculating to obtain a temperature difference, respectively inputting the temperature difference into a support vector regression model and the prediction model, and generating a corresponding predicted reactor cooler temperature difference value; the final temperature difference prediction value of the reactor cooler is obtained, and the service life curve of the reactor cooler is calculated in combination with the cooling efficiency; and determining a performance evaluation grade of the reactor cooler, and determining a maintenance time point and a maintenance strategy according to the performance evaluation grade. Through efficient data processing and feature extraction, the performance evaluation grade of the reactor cooler is determined, and the accuracy of performance determination of the reactor cooler is improved.

Description

Technical Field

[0001] The present application relates to the technical field of performance evaluation, and in particular to a method and device for performance evaluation and predictive maintenance of a reactor cooler. Background Art

[0002] As a key component in the power system, the stability and reliability of the reactor cooler's performance are crucial to the safe operation of the entire power grid. However, during long-term operation, the reactor cooler will experience performance degradation and potential failures due to continuous load and environmental factors, which directly affects the stability of the power system and the reliability of power supply. Summary of the invention

[0003] The present application aims to solve one of the technical problems in the related art at least to some extent.

[0004] To this end, the first objective of the present application is to propose a method for performance evaluation and predictive maintenance of a reactor cooler.

[0005] The second object of the present application is to provide a device.

[0006] The third objective of the present application is to provide an electronic device.

[0007] A fourth objective of the present application is to provide a computer-readable storage medium.

[0008] A fifth object of the present application is to provide a computer program product.

[0009] To achieve the above-mentioned purpose, the first embodiment of the present application proposes a method for performance evaluation and predictive maintenance of a reactor cooler, comprising:

[0010] Using sensors to collect historical data of the reactor cooler in real time, the historical data including a process variable data set and an operation variable data set;

[0011] De-noising the historical data, and screening the data in the operational variable data set according to the mutual information value between each variable in the operational variable data set and the target variable to obtain an input variable feature set;

[0012] Inputting the input variable feature set into the prediction model, and combining the real-time monitoring data to collaboratively predict the cooling efficiency and future energy consumption data, to obtain a performance degradation curve of the reactor cooler;

[0013] Calculate a first temperature difference and a second temperature difference according to the data in the process variable data set, input the first temperature difference and the second temperature difference into a support vector regression model and the prediction model respectively, and generate a corresponding predicted reactor cooler temperature difference value;

[0014] Integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating a reactor cooler service life curve in combination with the cooling efficiency of the performance degradation curve;

[0015] The performance evaluation level of the reactor cooler is determined in combination with the performance degradation curve and the service life curve, and the maintenance time point and maintenance strategy are determined according to the performance evaluation level.

[0016] Optionally, the process variable data set includes: operating temperature, ambient temperature, current, reactor thermal resistance, and the operation variable data set includes: flow rate, energy consumption data, operating status, and maintenance records;

[0017] The operating temperature and the ambient temperature belong to a first sub-data set of the process variable data set;

[0018] The current and the reactor thermal resistance belong to a second sub-data set of the process variable data set.

[0019] Optionally, the denoising the historical data and screening the data in the operational variable data set according to the mutual information value between each variable in the operational variable data set and the target variable to obtain the input variable feature set includes:

[0020] Performing wavelet transformation on the historical data through a wavelet basis function, and adjusting a scale factor and a translation factor in the wavelet basis function;

[0021] Calculating the mutual information value between the variables in the operating variable set and the target variable after wavelet transformation;

[0022] The variables whose mutual information values ​​are greater than a preset threshold are retained to construct the input variable feature set.

[0023] Optionally, the input variable feature set is input into a prediction model, and the cooling efficiency and future energy consumption data are collaboratively predicted in combination with real-time monitoring data to obtain a performance degradation curve of the reactor cooler, including:

[0024] Inputting the variable feature set into a prediction model to obtain predicted output variables, wherein the predicted output variables include predicted cooling efficiency and predicted energy consumption;

[0025] Calculating the overall mean of the predicted cooling efficiency and the predicted energy consumption;

[0026] Calculate the sum of squares between groups according to the number of observations and the mean at each time point and the overall mean, and calculate the mean square between groups according to the sum of squares between groups;

[0027] Calculating the intra-group sum of squares according to the mean values ​​at each time point and the predicted output variable, and calculating the intra-group mean square according to the intra-group sum of squares;

[0028] The corresponding significance level is queried in the test critical value table according to the inter-group mean square and the intra-group mean square, and the performance degradation curve is constructed according to the significance level, the predicted cooling efficiency and the predicted energy consumption.

[0029] Optionally, the calculating the first temperature difference and the second temperature difference according to the data in the process variable data set includes:

[0030] subtracting the ambient temperature from the operating temperature in the first sub-data set to obtain the first temperature difference;

[0031] The resistance loss is obtained according to the current in the second sub-data set and the resistance value of the reactor cooler, and the resistance loss is divided by the thermal resistance of the reactor to obtain the second temperature difference.

[0032] Optionally, the step of integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating a reactor cooler service life curve in combination with the cooling efficiency of the performance degradation curve, comprises:

[0033] Performing integrated learning on the predicted reactor cooler temperature difference value generated by the support vector regression model and the predicted reactor cooler temperature difference value generated by the prediction model to generate a final reactor cooler temperature difference prediction value;

[0034] According to the final reactor cooler temperature difference, predicted energy consumption and predicted cooling efficiency, the service life curve of the reactor cooler is calculated by Monteschinger's life law. The formula is:

[0035]

[0036] Wherein, T is the service life, A is a constant related to the insulation material grade, α is a preset constant, Tr is the final reactor cooler temperature difference value, y1 is the predicted cooling efficiency, and y2 is the predicted energy consumption.

[0037] Optionally, the performance evaluation levels include:

[0038] First degradation level: good performance, long remaining service life, no immediate maintenance required;

[0039] Secondary degradation level: stable performance, medium remaining service life, regular inspection is recommended;

[0040] Level 3 degradation level: performance begins to degrade, the remaining service life is short, and planned maintenance is recommended;

[0041] Level 4 degradation level: performance is significantly degraded, the remaining service life is very short, and maintenance is required immediately;

[0042] The maintenance time and maintenance strategy include:

[0043] Level 1 degradation: monitor performance, no maintenance required;

[0044] Level 2 degradation level: Arrange for inspection in the next operating cycle, including visual inspection and basic functional test;

[0045] Level 3 degradation: Develop a detailed maintenance plan including cleaning, parts replacement and performance testing;

[0046] Degradation level 4: Perform urgent maintenance measures immediately to avoid potential failures and downtime.

[0047] To achieve the above-mentioned purpose, the second embodiment of the present application proposes a reactor cooler performance evaluation and predictive maintenance device, comprising:

[0048] A collection module, used for collecting historical data of the reactor cooler in real time by using sensors, wherein the historical data includes a process variable data set and an operation variable data set;

[0049] A preprocessing module, used for performing denoising on the historical data, and screening the data in the operating variable data set according to the mutual information value between each variable in the operating variable data set and the target variable to obtain an input variable feature set;

[0050] A curve prediction module, used to input the input variable feature set into the prediction model, and to coordinately predict the cooling efficiency and future energy consumption data in combination with the real-time monitoring data to obtain a performance degradation curve of the reactor cooler;

[0051] a temperature difference prediction module, configured to calculate a first temperature difference and a second temperature difference according to data in the process variable data set, input the first temperature difference and the second temperature difference into a support vector regression model and the prediction model respectively, and generate a corresponding predicted reactor cooler temperature difference value;

[0052] A life prediction module, used for integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating a reactor cooler service life curve in combination with the cooling efficiency of the performance degradation curve;

[0053] The performance evaluation module is used to determine the performance evaluation level of the reactor cooler in combination with the performance degradation curve and the service life curve, and to determine the maintenance time point and maintenance strategy according to the performance evaluation level.

[0054] To achieve the above-mentioned purpose, the third aspect of the present application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0055] The memory stores computer-executable instructions;

[0056] The processor executes the computer-executable instructions stored in the memory to implement the method as described in any one of the first aspects.

[0057] To achieve the above-mentioned purpose, the fourth aspect embodiment of the present application proposes a computer-readable storage medium, in which computer-readable storage medium is stored computer execution instructions, and when the computer execution instructions are executed by a processor, they are used to implement the method as described in any one of the first aspects.

[0058] To achieve the above-mentioned purpose, the fifth aspect of the present application proposes a computer program product, which implements any method in the first aspect when executed by a processor.

[0059] The reactor cooler performance evaluation and predictive maintenance method, device, electronic device and storage medium provided in the present application realize the performance evaluation level determination of the reactor cooler through efficient data processing and feature extraction, and improve the accuracy of the reactor cooler performance determination.

[0060] Accurate performance degradation prediction: Through the Sageformer prediction model, combined with real-time monitoring data, this system can collaboratively predict cooling efficiency and energy consumption data, and use efficiency-to-energy ratio and variance analysis to generate performance degradation curves, providing a scientific basis for performance evaluation.

[0061] Accurate remaining service life estimation: The system uses parameters such as operating temperature, ambient temperature, current and reactor thermal resistance to predict the temperature difference through the SVR and Sageformer model integration, and uses Monte Schinger's life law to calculate the service life T curve, which improves the accuracy of service life prediction.

[0062] Real-time and early warning capabilities: The system can respond to changes in the operating status of the reactor cooler in real time, quickly diagnose performance degradation, and issue early warning signals in a timely manner, providing real-time support for maintenance decisions.

[0063] Customized maintenance suggestions: Based on the performance degradation curve and service life T curve, the system divides the evaluation levels into different levels and proposes corresponding maintenance time points and content suggestions, thus realizing the customization and optimization of maintenance activities.

[0064] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0066] Figure 1 A schematic flow chart of a reactor cooler performance evaluation and predictive maintenance method provided in an embodiment of the present application;

[0067] Figure 2 A schematic diagram of the structure of a reactor cooler performance evaluation and predictive maintenance device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0068] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0069] As a key component in the power system, the stability and reliability of the reactor cooler's performance are crucial to the safe operation of the entire power grid. However, during long-term operation, the reactor cooler will experience performance degradation and potential failures due to continuous load and environmental factors, which directly affects the stability of the power system and the reliability of power supply. Therefore, developing an effective reactor cooler performance evaluation and predictive maintenance system is of great significance for identifying performance degradation trends in advance, preventing failures, and optimizing maintenance plans.

[0070] At present, the performance evaluation and predictive maintenance methods of reactor coolers mainly rely on traditional time or periodic maintenance strategies, which are often based on experience and lack in-depth analysis and prediction of the real-time operating status of the equipment. In addition, although some existing predictive maintenance methods utilize sensor data and statistical analysis techniques, most of them ignore the complexity and nonlinear characteristics of equipment operation data, making it difficult to achieve early identification and accurate prediction of performance degradation. At the same time, these methods also face challenges such as insufficient feature extraction, limited model generalization ability, and insufficient real-time performance when processing large amounts of multi-dimensional and high-noise data generated in the actual operation of the equipment.

[0071] The main problems of existing reactor cooler performance evaluation and predictive maintenance methods include: Limited data analysis capabilities: Traditional methods are difficult to effectively process and analyze the large amount of multidimensional data generated during the operation of reactor coolers. Low prediction accuracy: The accuracy and reliability of existing models in predicting performance degradation trends still need to be improved. Insufficient real-time performance: The speed of performance evaluation often cannot keep up with the needs of real-time monitoring. Insufficient maintenance strategy optimization: There is a lack of customized and optimized maintenance strategies based on performance evaluation results.

[0072] To address this problem, the present invention provides a method for evaluating and predicting the performance of a reactor cooler. Figure 1 A flow chart of a reactor cooler performance evaluation and predictive maintenance method provided in an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0073] Step 101, using sensors to collect historical data of a reactor cooler in real time, wherein the historical data includes a process variable data set and an operation variable data set;

[0074] Step 102, performing denoising processing on the historical data, and screening the data in the operating variable data set according to the mutual information value between each variable in the operating variable data set and the target variable to obtain an input variable feature set;

[0075] Step 103, inputting the input variable feature set into the prediction model, and combining the real-time monitoring data to collaboratively predict the cooling efficiency and future energy consumption data, to obtain a performance degradation curve of the reactor cooler;

[0076] Step 104, calculating a first temperature difference and a second temperature difference according to the data in the process variable data set, inputting the first temperature difference and the second temperature difference into a support vector regression model and the prediction model respectively to generate a corresponding predicted reactor cooler temperature difference value;

[0077] Step 105, integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating a reactor cooler service life curve in combination with the cooling efficiency of the performance degradation curve;

[0078] Step 106: Determine a performance evaluation level of the reactor cooler in combination with the performance degradation curve and the service life curve, and determine a maintenance time point and a maintenance strategy according to the performance evaluation level.

[0079] Optionally, the process variable data set includes: operating temperature, ambient temperature, current, reactor thermal resistance, and the operation variable data set includes: flow rate, energy consumption data, operating status, and maintenance records;

[0080] The operating temperature and the ambient temperature belong to a first sub-data set of the process variable data set;

[0081] The current and the reactor thermal resistance belong to the second sub-data set of the process variable data set.

[0082] In this embodiment, the operating temperature and the ambient temperature are used to construct a data set S1 = (V1, V2), the current and the reactor thermal resistance are used to construct a data set S2 = (I, Φ), and the operation variables are used to construct a data set S3 = (x1, ..., x n ), where n is the number of operation variable data sets.

[0083] Optionally, the step 102 performs denoising on the historical data, and screens the data in the operational variable data set according to the mutual information value between each variable in the operational variable data set and the target variable to obtain an input variable feature set, including:

[0084] Performing wavelet transformation on the historical data through a wavelet basis function, and adjusting a scale factor and a translation factor in the wavelet basis function;

[0085] Calculating the mutual information value between the variables in the operating variable set and the target variable after wavelet transformation;

[0086] The variables whose mutual information values ​​are greater than a preset threshold are retained to construct the input variable feature set.

[0087] In this embodiment, the wavelet transform algorithm is suitable for analyzing non-stationary signals, suitable for capturing instantaneous events and local characteristics of signals, and is convenient for understanding and analyzing signal characteristics. The specific implementation process is as follows:

[0088] f(s) is the signal of the process variable data set and the operation variable data set, and its expression is:

[0089]

[0090] Among them, Ψ(s) is the wavelet basis, is the conjugate function of Ψ(s).

[0091] In the process of wavelet transform, wavelet basis function needs to be stretched and translated to cooperate with signal analysis. After the translation transformation of wavelet basis function, we have:

[0092]

[0093] Among them, a represents the scale factor, which affects the waveform width of Ψ(s), a≠0; b represents the translation factor, a,b∈R.

[0094] Wavelet transform has the characteristics of multi-resolution analysis. By scaling and translating the wavelet basis function according to the scale factor a and the translation factor b, wavelet bases of different widths can be obtained to match any local time domain position of the signal, thereby achieving the purpose of obtaining the local time domain information of the signal.

[0095] The expression of wavelet reconstructed signal (inverse transform) is as follows:

[0096]

[0097] Among them, Ψf(a,b) is the coefficient of wavelet transform, that is, f(s) and wavelet basis function Ψ a,b (s), C is a constant.

[0098] In this embodiment, the mutual information value (MI) is a measure of the mutual dependence between two random variables, which can be used to measure the correlation between the feature and the target variable. The basic method for calculating the mutual information value is as follows:

[0099]

[0100] Where I(X;Y) is the mutual information value between X and Y, p(x,y) and p(x,y) are the joint probability density functions of X and Y, and p(x) and p(y) are the marginal probability density functions of X and Y, respectively.

[0101] The larger the mutual information value, the greater the contribution of the feature to the target variable and the stronger its predictive ability. By selecting the feature with the highest mutual information value, the most relevant features to the target variable can be extracted, thereby improving the performance of the machine learning model.

[0102] The set of operating variables after wavelet transformation is recorded as S3'=(x1,...,x n ) Use mutual information MI for feature selection. Mutual information measures the interdependence between two variables and can be used to identify the features that are most relevant to the output variable. The steps for multivariate feature selection using mutual information are as follows:

[0103] The input variable is S3', and the output variable is cooling efficiency and energy consumption, recorded as Y = (y1, y2). The mutual information value MI (S3', Y) between them is calculated as follows:

[0104]

[0105] Among them, P(x,y) is the joint probability that S3' takes the value x and the output variable Y takes the value y, and P(x) and P(y) are the marginal probabilities of x and y respectively.

[0106] The input variables are sorted according to the mutual information value. The larger the mutual information, the stronger the dependence between the input variable and the output variable. The threshold is set to select the top-ranked input variables to construct the input variable feature set S = (x1, x3, ..., x n-3 ) is input into the prediction model to predict cooling efficiency and future energy consumption.

[0107] Optionally, the step 103 inputs the input variable feature set into a prediction model, and combines the real-time monitoring data to collaboratively predict cooling efficiency and future energy consumption data to obtain a performance degradation curve of the reactor cooler, including:

[0108] Inputting the variable feature set into a prediction model to obtain predicted output variables, wherein the predicted output variables include predicted cooling efficiency and predicted energy consumption;

[0109] Calculating the overall mean of the predicted cooling efficiency and the predicted energy consumption;

[0110] Calculate the sum of squares between groups according to the number of observations and the mean at each time point and the overall mean, and calculate the mean square between groups according to the sum of squares between groups;

[0111] Calculating the intra-group sum of squares according to the mean values ​​at each time point and the predicted output variable, and calculating the intra-group mean square according to the intra-group sum of squares;

[0112] The corresponding significance level is queried in the test critical value table according to the inter-group mean square and the intra-group mean square, and the performance degradation curve is constructed according to the significance level, the predicted cooling efficiency and the predicted energy consumption.

[0113] In this embodiment, the prediction model is a SageFormer model, including:

[0114] ① Input layer: receives input variable feature set S = (x1, x3, ..., x n-3 ) and real-time monitoring data, dividing it into multiple small blocks, which not only improves the computational efficiency and memory usage efficiency of the model, but also enhances the ability to capture local patterns, and maintains the learning of long-distance dependencies through parallel processing and self-attention mechanisms, thereby providing flexibility and adaptability for the SageFormer model in long-term multivariate time series prediction tasks.

[0115] ②Graph structure learning layer: It is equivalent to a special "hidden layer" where the learned graph structure is used to represent the complex relationship between sequences. This layer uses a graph neural network (GNN) to capture the dependencies between sequences.

[0116] ③ Sequence-aware global Token layer: As a bridge between the input layer and the graph structure layer, the global Token is used to capture the global statistical characteristics of each sequence and enhance the interaction between sequences through the graph structure layer.

[0117] ④ Iterative message passing layer: This is the core part of the model. Through the iterative process of graph aggregation and time encoding, the representation of the sequence is continuously updated to fuse information from other sequences and capture temporal dependencies.

[0118] ⑤Transformer Encoder Block (TEB): As part of the hidden layer, it uses the self-attention mechanism to handle the temporal dependencies within the sequence and further integrates the information obtained from the graph aggregation layer.

[0119] ⑥ Output layer: The advantage of using the prediction head is that it simplifies the linear decoder, reduces the complexity and computational cost of the model, and improves the efficiency and accuracy of the prediction.

[0120] The variance analysis method was used to evaluate the significant difference between the cooling efficiency and future energy consumption predicted by the model. The steps are as follows:

[0121] Calculate the overall average of cooling efficiency and future energy consumption:

[0122]

[0123] Where N is the total number of output variables and k represents the kth output variable.

[0124] Calculate the between-group sum of squares SSB:

[0125]

[0126] Where m is the number of time points, u j is the number of observations at the jth time point, is the mean value at the jth time point.

[0127] Calculate the within-group sum of squares SSW:

[0128]

[0129] Among them, Y ij k is the kth output variable, the i-th observation at the j-th time point.

[0130] Calculate the between-group and within-group mean squares (MSB and MSW):

[0131]

[0132] Calculate the F statistic:

[0133]

[0134] Determine the significance level, and determine the p-value by looking up the F distribution table or using statistical software; when the p-value is less than the significance level, it is considered that the mean of the cooling efficiency or energy consumption at at least one time point is significantly different from that at other time points, and the performance degradation curve is constructed in conjunction with the efficiency-energy consumption ratio:

[0135]

[0136] Optionally, the step 104 of calculating the first temperature difference and the second temperature difference according to the data in the process variable data set includes:

[0137] subtracting the ambient temperature from the operating temperature in the first sub-data set to obtain the first temperature difference;

[0138] The resistance loss is obtained according to the current in the second sub-data set and the resistance value of the reactor cooler, and the resistance loss is divided by the thermal resistance of the reactor to obtain the second temperature difference.

[0139] In this embodiment, the above-constructed S1=(V1, V2) and constructed S2=(I, Φ) are subjected to wavelet transformation to obtain data sets S1'=(V1', V2') and S2'=(I', Φ'), and the temperature difference of the inductive cooler is calculated respectively.

[0140] The formula for calculating the first temperature difference input by the data set S1' is as follows:

[0141] Tr1=V1′-V2′ (12)

[0142] The second temperature difference is calculated by using the data set S2'. First, the resistance loss is calculated: P = (I') 2 R, where R is the resistance of the reactor cooler; the loss Q in the reactor is mainly dissipated in the form of heat, so the heat generated can be calculated by the resistance loss, which is directly related to the resistance loss P, that is, Q = P; the thermal resistance Φ is a parameter that describes the difficulty of heat transfer from the heat source to the surrounding environment, and the second temperature difference formula is as follows:

[0143]

[0144] Since the calculation process of the second temperature difference is complicated, the above-mentioned Sageformer model is used to predict the future reactor cooler temperature difference 2; the first temperature difference uses the support vector regression (SVR) model to predict the future reactor group cooler temperature difference, where the SVR model process is as follows:

[0145] ① The input layer is used to receive an input variable set 1, which includes a plurality of first temperature differences;

[0146] ② The feature mapping layer maps the input variable set 1 to a high-dimensional space through a mapping function;

[0147] ③SVR uses the kernel function to calculate the similarity between two points in the original feature space or high-dimensional space. Its expression is as follows:

[0148]

[0149] Among them, (β i -β i * ) represents the Lagrange multiplier of the dual space transformation, O(Tr1,Tr1 i ) represents the kernel function used to solve the quadratic equation, and b represents the bias.

[0150] ④The output layer outputs the future temperature difference of the reactance group cooler.

[0151] Optionally, the step 105 integrates the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculates the service life curve of the reactor cooler in combination with the cooling efficiency of the performance degradation curve, including:

[0152] Performing integrated learning on the predicted reactor cooler temperature difference value generated by the support vector regression model and the predicted reactor cooler temperature difference value generated by the prediction model to generate a final reactor cooler temperature difference prediction value;

[0153] According to the final reactor cooler temperature difference, predicted energy consumption and predicted cooling efficiency, the service life curve of the reactor cooler is calculated by Monteschinger's life law. The formula is:

[0154]

[0155] Wherein, T is the service life, A is a constant related to the insulation material grade, α is a preset constant, Tr is the final reactor cooler temperature difference value, y1 is the predicted cooling efficiency, and y2 is the predicted energy consumption.

[0156] The predicted data is integrated through Blanding to obtain the final predicted value of reactor cooler temperature difference. Blending ensemble learning mainly includes meta-learning layer and generalized model layer (GM layer).

[0157] The core function of the meta-learning layer is to use the prediction results of multiple basic models as input features and make the final decision by learning the relationship between these predictions and actual outputs. This method not only improves the generalization ability and prediction accuracy of the model, but also reduces the risk of overfitting and the possibility of information leakage.

[0158] The GM layer uses the prediction results of the meta-learning layer as input features to construct the final prediction results. The role of this layer is to optimize and adjust the prediction weights of each basic model to improve the accuracy and generalization ability of the overall prediction. The design of the GM layer allows it to capture the relationship between different models and reduce uncertainty and potential bias through an integrated approach, thereby providing more robust performance in complex prediction tasks.

[0159] Based on blending ensemble learning, the prediction results of the future reactor group cooler temperature difference value 1 and the future reactor group cooler temperature difference value 2 are integrated to obtain the final reactor cooler temperature difference value Tr.

[0160] The final reactor cooler temperature difference value is combined with the cooling efficiency predicted in the predicted performance degradation curve module to calculate the service life of the reactor cooler insulation material through the Monteschinger life law, that is, the reactor cooler service life T curve. The formula is as follows:

[0161]

[0162] Among them, A is a constant related to the insulation material grade, α is a constant generally 0.88, Tr is the final reactor cooler temperature difference, is the efficiency / energy consumption ratio.

[0163] Optionally, the performance evaluation levels include:

[0164] First degradation level: good performance, long remaining service life, no immediate maintenance required;

[0165] Secondary degradation level: stable performance, medium remaining service life, regular inspection is recommended;

[0166] Level 3 degradation level: performance begins to degrade, the remaining service life is short, and planned maintenance is recommended;

[0167] Level 4 degradation level: performance is significantly degraded, the remaining service life is very short, and maintenance is required immediately;

[0168] The maintenance time and maintenance strategy include:

[0169] Level 1 degradation: monitor performance, no maintenance required;

[0170] Level 2 degradation level: Arrange for inspection in the next operating cycle, including visual inspection and basic functional test;

[0171] Level 3 degradation: Develop a detailed maintenance plan including cleaning, parts replacement and performance testing;

[0172] Degradation level 4: Perform urgent maintenance measures immediately to avoid potential failures and downtime.

[0173] The present invention proposes a reactor cooler performance evaluation and predictive maintenance system, which aims to optimize the operation monitoring and maintenance strategy of the reactor cooler. The system consists of three key modules: performance degradation prediction, remaining service life estimation and evaluation warning. In the performance degradation prediction module, the system collects the operating parameters of the cooler in real time through sensors, such as temperature, flow rate, energy consumption, etc., and preprocesses and extracts features of the data through wavelet transform and mutual information algorithm. The Sageformer model is used to combine real-time data to predict the cooling efficiency and energy consumption, and then generate a performance degradation curve. In the remaining service life estimation module, the system calculates the operating temperature difference of the cooler, and uses SVR and Sageformer models for prediction, and obtains the final temperature difference value through the integrated method Blending. Combined with the cooling efficiency, according to the Monteschinger life law, the expected service life of the reactor cooler is calculated. Finally, the evaluation and warning module divides the performance of the cooler into four levels according to the performance degradation curve and the service life T curve, and puts forward targeted maintenance suggestions accordingly to achieve preventive maintenance and improve the operating reliability and maintenance efficiency of the reactor cooler.

[0174] In order to implement the above-mentioned embodiment, the present application also proposes a reactor cooler performance evaluation and predictive maintenance device. Figure 2 A schematic diagram of the structure of a reactor cooler performance evaluation and predictive maintenance device provided in an embodiment of the present application. Figure 2 As shown, the device comprises:

[0175] A collection module, used for collecting historical data of the reactor cooler in real time by using sensors, wherein the historical data includes a process variable data set and an operation variable data set;

[0176] A preprocessing module, used for performing denoising on the historical data, and screening the data in the operating variable data set according to the mutual information value between each variable in the operating variable data set and the target variable to obtain an input variable feature set;

[0177] A curve prediction module, used to input the input variable feature set into the prediction model, and to coordinately predict the cooling efficiency and future energy consumption data in combination with the real-time monitoring data to obtain a performance degradation curve of the reactor cooler;

[0178] a temperature difference prediction module, configured to calculate a first temperature difference and a second temperature difference according to data in the process variable data set, input the first temperature difference and the second temperature difference into a support vector regression model and the prediction model respectively, and generate a corresponding predicted reactor cooler temperature difference value;

[0179] A life prediction module, used for integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating a reactor cooler service life curve in combination with the cooling efficiency of the performance degradation curve;

[0180] The performance evaluation module is used to determine the performance evaluation level of the reactor cooler in combination with the performance degradation curve and the service life curve, and to determine the maintenance time point and maintenance strategy according to the performance evaluation level.

[0181] In order to implement the above embodiments, the present application also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method provided by the above embodiments.

[0182] In order to implement the above embodiments, the present application also proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0183] In order to implement the above embodiments, the present application also proposes a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0184] The collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in this application are in compliance with relevant laws and regulations and do not violate public order and good morals.

[0185] It should be noted that personal information from users should be collected for legitimate and reasonable purposes and should not be shared or sold outside of these legitimate uses. In addition, such collection / sharing should be carried out after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign the agreement / authorization including authorization of relevant user information before the user uses the function. In addition, any necessary steps should be taken to protect and safeguard access to such personal information data and ensure that others who have access to personal information data comply with its privacy policy and procedures.

[0186] The present application is expected to provide an implementation scheme for users to selectively block the use or access of personal information data. That is, the present disclosure is expected to provide hardware and / or software to prevent or block access to such personal information data. Once the personal information data is no longer needed, the risk can be minimized by limiting data collection and deleting the data. In addition, when applicable, such personal information is de-identified to protect the privacy of the user.

[0187] In the description of the aforementioned embodiments, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0188] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0189] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0190] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0191] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0192] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0193] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0194] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for performance evaluation and predictive maintenance of a reactor cooler, characterized in that: The following steps are involved: Using sensors to collect historical data of the reactor cooler in real time, the historical data including a process variable data set and an operation variable data set; De-noising the historical data, and screening the data in the operational variable data set according to the mutual information value between each variable in the operational variable data set and the target variable to obtain an input variable feature set; Inputting the input variable feature set into the prediction model, and combining the real-time monitoring data to collaboratively predict the cooling efficiency and future energy consumption data, to obtain a performance degradation curve of the reactor cooler; Calculate a first temperature difference and a second temperature difference according to the data in the process variable data set, input the first temperature difference and the second temperature difference into a support vector regression model and the prediction model respectively, and generate a corresponding predicted reactor cooler temperature difference value; Integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating a reactor cooler service life curve in combination with the cooling efficiency of the performance degradation curve; The performance evaluation level of the reactor cooler is determined in combination with the performance degradation curve and the service life curve, and the maintenance time point and maintenance strategy are determined according to the performance evaluation level.

2. The method according to claim 1, characterized in that The process variable data set includes: operating temperature, ambient temperature, current, reactor thermal resistance, and the operation variable data set includes: flow rate, energy consumption data, operating status, and maintenance records; The operating temperature and the ambient temperature belong to a first sub-data set of the process variable data set; The current and the reactor thermal resistance belong to the second sub-data set of the process variable data set.

3. The method according to claim 2, characterized in that The denoising process is performed on the historical data, and the data in the operation variable data set is screened according to the mutual information value between each variable in the operation variable data set and the target variable to obtain the input variable feature set, including: Performing wavelet transformation on the historical data through a wavelet basis function, and adjusting a scale factor and a translation factor in the wavelet basis function; Calculating the mutual information value between the variables in the operating variable set and the target variable after wavelet transformation; The variables whose mutual information values ​​are greater than a preset threshold are retained to construct the input variable feature set.

4. The method according to claim 3, characterized in that The input variable feature set is input into the prediction model, and the cooling efficiency and future energy consumption data are predicted in coordination with the real-time monitoring data to obtain the performance degradation curve of the reactor cooler, including: Inputting the variable feature set into a prediction model to obtain predicted output variables, wherein the predicted output variables include predicted cooling efficiency and predicted energy consumption; Calculating the overall mean of the predicted cooling efficiency and the predicted energy consumption; Calculate the sum of squares between groups according to the number of observations and the mean at each time point and the overall mean, and calculate the mean square between groups according to the sum of squares between groups; Calculating the intra-group sum of squares according to the mean values ​​at each time point and the predicted output variable, and calculating the intra-group mean square according to the intra-group sum of squares; The corresponding significance level is queried in the test critical value table according to the inter-group mean square and the intra-group mean square, and the performance degradation curve is constructed according to the significance level, the predicted cooling efficiency and the predicted energy consumption.

5. The method according to claim 4, characterized in that The step of calculating the first temperature difference and the second temperature difference according to the data in the process variable data set comprises: subtracting the ambient temperature from the operating temperature in the first sub-data set to obtain the first temperature difference; The resistance loss is obtained according to the current in the second sub-data set and the resistance value of the reactor cooler, and the resistance loss is divided by the thermal resistance of the reactor to obtain the second temperature difference.

6. The method according to claim 5, characterized in that The step of integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating the service life curve of the reactor cooler in combination with the cooling efficiency of the performance degradation curve includes: Performing integrated learning on the predicted reactor cooler temperature difference value generated by the support vector regression model and the predicted reactor cooler temperature difference value generated by the prediction model to generate a final reactor cooler temperature difference prediction value; According to the final reactor cooler temperature difference, predicted energy consumption and predicted cooling efficiency, the service life curve of the reactor cooler is calculated by Monteschinger's life law. The formula is: Wherein, T is the service life, A is a constant related to the insulation material grade, α is a preset constant, Tr is the final reactor cooler temperature difference value, y1 is the predicted cooling efficiency, and y2 is the predicted energy consumption.

7. The method according to claim 6, characterized in that The performance evaluation levels include: First degradation level: good performance, long remaining service life, no immediate maintenance required; Secondary degradation level: stable performance, medium remaining service life, regular inspection is recommended; Level 3 degradation level: performance begins to degrade, the remaining service life is short, and planned maintenance is recommended; Level 4 degradation level: performance is significantly degraded, the remaining service life is very short, and maintenance is required immediately; The maintenance time and maintenance strategy include: Level 1 degradation: monitor performance, no maintenance required; Level 2 degradation level: Arrange for inspection in the next operating cycle, including visual inspection and basic functional test; Level 3 degradation: Develop a detailed maintenance plan including cleaning, parts replacement and performance testing; Degradation level 4: Perform urgent maintenance measures immediately to avoid potential failures and downtime.

8. A reactor cooler performance evaluation and predictive maintenance device, characterized in that: include: A collection module, used for collecting historical data of the reactor cooler in real time by using sensors, wherein the historical data includes a process variable data set and an operation variable data set; A preprocessing module, used for performing denoising on the historical data, and screening the data in the operating variable data set according to the mutual information value between each variable in the operating variable data set and the target variable to obtain an input variable feature set; A curve prediction module, used to input the input variable feature set into the prediction model, and to coordinately predict the cooling efficiency and future energy consumption data in combination with the real-time monitoring data to obtain a performance degradation curve of the reactor cooler; a temperature difference prediction module, configured to calculate a first temperature difference and a second temperature difference according to data in the process variable data set, input the first temperature difference and the second temperature difference into a support vector regression model and the prediction model respectively, and generate a corresponding predicted reactor cooler temperature difference value; A life prediction module, used for integrating the predicted reactor cooler temperature difference to obtain a final reactor cooler temperature difference prediction value, and calculating a reactor cooler service life curve in combination with the cooling efficiency of the performance degradation curve; The performance evaluation module is used to determine the performance evaluation level of the reactor cooler in combination with the performance degradation curve and the service life curve, and to determine the maintenance time point and maintenance strategy according to the performance evaluation level.

9. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images