Fault prediction method and system for circulating water system of air compressor, electronic equipment and storage medium

Abstract

The invention relates to the technical field of air compressors, and particularly provides an air compressor circulating water system fault prediction method and system, electronic equipment and a storage medium, and the method comprises the steps: obtaining historical operation data of an air compressor circulating water system, and carrying out the preprocessing; extracting first statistical characteristics of water quality indexes and material corrosion rates in the historical operation data according to a preset time window; extracting a second statistical characteristic of the cooling efficiency in combination with the working load of the air compressor; constructing and training a fault prediction model based on the first statistical feature and the second statistical feature, and outputting a fault prediction result by using the trained fault prediction model; based on the fault prediction results, a dynamic maintenance strategy is generated with minimized downtime and maintenance cost. The objective of the invention is to solve the technical problem of insufficient fault prediction accuracy of an air compressor circulating water system and improve the cooling efficiency of an air compressor cooler.

Description

Technical Field

[0001] The present invention relates to the technical field of air compressors, and in particular, to a method, a system, an electronic device, and a storage medium for predicting faults in the circulating water system of an air compressor. Background Art

[0002] In an air compressor system, the cooler plays a crucial role. It improves the efficiency and reliability of the system by reducing the temperature of the compressed air. Usually, water-cooling technology is adopted, using circulating cooling water as the heat exchange medium to absorb the heat generated by the compression process and transfer this heat to the external environment. However, during the actual operation process, the quality of the circulating cooling water has a direct impact on the performance of the entire cooling system.

[0003] Currently, the maintenance of the circulating water system of an air compressor is mostly based on fixed cycles or relies on the experience of operators, lacking effective real-time monitoring means. The maintenance mode based on time intervals cannot timely capture the changing trends under actual working conditions, making it difficult to accurately predict faults in the circulating water system of an air compressor; over-maintenance will cause waste of resources, while insufficient maintenance may increase the risk of equipment damage, and ultimately both are reflected in the decrease of the cooling efficiency of the air compressor cooler. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, a system, an electronic device, and a storage medium for predicting faults in the circulating water system of an air compressor, so as to solve the technical problem of insufficient accuracy in predicting faults in the circulating water system of an air compressor and improve the cooling efficiency of the air compressor cooler.

[0005] The technical solution of the first aspect of the present invention provides a method for predicting faults in the circulating water system of an air compressor, and the method includes:

[0006] Obtain the historical operation data of the circulating water system of the air compressor and perform preprocessing;

[0007] Extract the first statistical features of water quality indicators and material corrosion rates from the historical operation data according to a preset time window;

[0008] Combine the working load of the air compressor to extract the second statistical features of the cooling efficiency;

[0009] Construct and train a fault prediction model based on the first statistical features and the second statistical features, and use the trained fault prediction model to output a fault prediction result;

[0010] Based on the fault prediction result, generate a dynamic maintenance strategy to minimize the downtime and maintenance cost.

[0011] Furthermore, the historical operation data of the circulating water system of the air compressor at least includes water quality index data, material corrosion data, inlet and outlet temperature data, and working load data.

[0012] Further, the first statistical features of water quality indicators and material corrosion rate extracted from historical operation data according to a preset time window include:

[0013] Extract the water quality indicator vector and the material corrosion rate vector from the historical operation data according to the preset time window respectively;

[0014] Use the water quality indicator vector and the material corrosion rate vector to extract the first statistical features, including: for each component in the water quality indicators and the material corrosion rate, calculate the rolling average of the water quality indicators and the rolling average of the material corrosion rate respectively; extract the rolling standard deviation of the water quality indicators and the rolling standard deviation of the material corrosion rate respectively based on the rolling average of the water quality indicators and the rolling average of the material corrosion rate.

[0015] Further, the second statistical features of the cooling efficiency extracted in combination with the working load of the air compressor include:

[0016] Extract the change rate of the cooling efficiency based on the working load of the air compressor circulating water system at different times;

[0017] Calculate the weighted rolling average and standard deviation of the cooling efficiency change rate.

[0018] Further, construct and train a fault prediction model based on the first statistical features and the second statistical features, and use the trained fault prediction model to output the fault prediction result, including:

[0019] Use the LSTM network to extract the time series feature representation based on the first statistical features;

[0020] Use the XGBoost model to construct a fault prediction model, train the fault prediction model based on the time series feature representation and the second statistical features, and use the trained fault prediction model to output the fault prediction result.

[0021] Further, using the LSTM network to extract the time series feature representation based on the first statistical features includes:

[0022] Generate a time series data set based on the rolling average and rolling standard deviation of the water quality indicators and the rolling average and rolling standard deviation of the material corrosion rate;

[0023] At each time step, the long short-term memory network receives the input vector at the current time step and the state information at the previous time step, and updates the internal state through the gating mechanism;

[0024] After the LSTM processes the input vectors at all times, output a vector containing the information of the entire sequence and use this vector as the time series feature representation of the air compressor circulating water system.

[0025] Further, an XGBoost model is used to construct a fault prediction model. The fault prediction model is trained based on the time series feature representation and the second statistical feature. The fault prediction results output by the trained fault prediction model include:

[0026] Combine the time series feature representation and the second statistical feature;

[0027] With the goal of maximizing the gain, train the fault prediction model based on the combined time series feature representation and the second statistical feature;

[0028] Output the fault prediction results using the trained fault prediction model.

[0029] The technical solution of the second aspect of the present invention provides a fault prediction system for an air compressor circulating water system. The fault prediction system includes:

[0030] A data acquisition module configured to obtain the historical operation data of the air compressor circulating water system and perform preprocessing;

[0031] A first statistical feature extraction module configured to extract the first statistical features of water quality indicators and material corrosion rate from the historical operation data according to a preset time window

[0032] A second statistical feature extraction module configured to extract the second statistical features of cooling efficiency in combination with the working load of the air compressor

[0033] A fault prediction module configured to construct and train a fault prediction model based on the first statistical feature and the second statistical feature, and output the fault prediction results using the trained fault prediction model;

[0034] A strategy generation module configured to generate a dynamic maintenance strategy based on the fault prediction results with the goal of minimizing downtime and maintenance costs.

[0035] The technical solution of the third aspect of the present invention provides an electronic device. The electronic device includes: a processor and a memory communicatively connected to the processor; wherein, the memory stores instructions executable by the processor, and when the instructions are executed by the processor, the processor can execute the steps of the method for predicting faults in the air compressor circulating water system according to the technical solution of the first aspect of the present invention.

[0036] The technical solution of the fourth aspect of the present invention provides a computer-readable storage medium. A program for implementing the method for predicting faults in the air compressor circulating water system is stored on the computer-readable storage medium. When the program for implementing the method for predicting faults in the air compressor circulating water system is executed by a processor, the steps of the method for predicting faults in the air compressor circulating water system according to the technical solution of the first aspect of the present invention are implemented.

[0037] The technical solutions of the embodiments of the present invention have at least the following advantages and beneficial effects:

[0038] The method for predicting faults in the air compressor circulating water system provided by the present invention constructs a training fault detection model by combining the statistical characteristics of water quality indicators, material corrosion rate, and cooling efficiency. The water quality indicators, material corrosion rate, and cooling efficiency respectively reflect the operating state of the system from different perspectives, effectively reflecting the change trend of the system. The fault detection model, based on a more comprehensive system view, comprehensively considers the statistical characteristics of water quality indicators, material corrosion rate, and cooling efficiency, captures their complex interactions, and thus predicts faults more accurately, improving the accuracy of fault prediction for the air compressor circulating water system. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments of the present invention. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0040] Figure 1 It is a schematic diagram of the method for predicting faults in the air compressor circulating water system provided by the embodiments of the present invention;

[0041] Figure 2 It is a schematic diagram of the structure of the air compressor circulating water system fault prediction system provided by the embodiments of the present invention;

[0042] Figure 3 It is a schematic diagram of the structure of the air compressor circulating water system provided by the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated in the drawings here can be arranged and designed in various different configurations.

[0044] It should be understood that the "system", "device", "unit", and / or "module" used herein is a method for distinguishing different components, elements, parts, portions, or assemblies at different levels. However, if other words can achieve the same purpose, the described words can be replaced by other expressions.

[0045] As shown in this specification and the claims, unless the context clearly indicates otherwise, words such as "a", "an", "one", and / or "the" are not specifically singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. A method or device may also include other steps or elements.

[0046] Flowcharts are used in this specification to illustrate the operations performed by the system according to the embodiments of this specification. It should be understood that the previous or subsequent operations do not necessarily need to be executed precisely in sequence. On the contrary, the steps can be processed in reverse order or simultaneously. At the same time, other operations can also be added to these processes, or one or more steps can be removed from these processes.

[0047] Please refer to Figure 1 As shown, the technical solution of the first aspect of the present invention provides a method for predicting faults in the air compressor circulating water system. The method includes:

[0048] Step S100: Obtain the historical operation data of the air compressor circulating water system and perform preprocessing; specifically, the air compressor circulating water system includes a water pump, a water tank, a cooling tower, an air compressor cooler, and a solenoid valve; the historical operation data of the air compressor circulating water system at least includes water quality index data, material corrosion data, inlet and outlet temperature data, and workload data; among them, the water quality index data at least includes pH value, conductivity, and turbidity; the material corrosion data is the corrosion data of metal components, specifically the corrosion data of the air compressor cooler. Specifically, a corrosion resistance probe is used to collect the corrosion data of the air compressor cooler. The corrosion resistance probe can provide real-time corrosion current density data, and then evaluate the corrosion degree of the material; data preprocessing at least includes data cleaning and standardization processing;

[0049] Step S200: Extract the statistical features of water quality indicators and material corrosion rates from the historical operation data according to a preset time window; specifically, the preset time window can be selected according to actual needs, preferably 24 hours or 7 days;

[0050] Step S200 specifically includes:

[0051] Step S210: Extract a water quality index vector and a material corrosion rate vector from the historical operation data according to the preset time window; specifically, the water quality index vector includes: pH value (PH): reflecting the acidity and alkalinity of water; conductivity (EC): reflecting the ion concentration in water; turbidity: reflecting the suspended solid content in water; the material corrosion rate vector is the corrosion rate of the air compressor cooler;

[0052] Step S220: Extract the first statistical features using the water quality index vector and the material corrosion rate vector;

[0053] Step S220 specifically includes:

[0054] Step S221: For each component of the water quality index and the material corrosion rate, calculate the rolling average of the water quality index and the rolling average of the material corrosion rate respectively, which can be expressed as:

[0055]

[0056]

[0057] In the formula, represents the rolling average value of the i-th water quality index at time t, i ∈ {PH, EC, Turbidity}; n represents the preset rolling window size; j represents the index variable; represents the rolling average value of the material corrosion rate at time t;

[0058] Step S222: Based on the rolling average of the water quality index and the rolling average of the material corrosion rate, extract the rolling standard deviation of the water quality index and the rolling standard deviation of the material corrosion rate respectively, which can be expressed as:

[0059]

[0060] In the formula, σw i (t) represents the rolling standard deviation of the i-th water quality index at time t; σc(t) represents the rolling standard deviation of the material corrosion rate at time t; By extracting the rolling standard deviations of the water quality index and the material corrosion rate, the volatility of these parameters over time can be obtained. These features reflect the average level and stability of the water quality and corrosion conditions within a specific time window. Using the first statistical features as the input of the fault prediction model can provide more stable and representative information, thereby improving the accuracy of the fault prediction model and discovering potential fault signs;

[0061] Step S300: Combine the working load of the air compressor to extract the second statistical feature of the cooling efficiency;

[0062] Step S300 specifically includes:

[0063] Step S310: Based on the working loads of the air compressor circulating water system at different times, extract the change rate of the cooling efficiency; Specifically, obtain the inlet temperature T in (t), the outlet temperature T out (t) and the working load L(t) of the air compressor circulating water system at each time point, and calculate the cooling efficiency and the change rate of the cooling efficiency. Among them, the cooling efficiency can be expressed as: E(t) = T in (t) - Tout (t); the cooling efficiency change rate can be expressed as: ΔE(t) = E(t) - E(t - 1); in this embodiment, a weighted function of the workload is also defined, which can be expressed as: q(t) = a1L(t) + b, where q(t) represents the weighted function of the workload; a1 represents a preset weight coefficient; b represents a bias term.

[0064] Step S320: Calculate the weighted rolling average and standard deviation of the cooling efficiency change rate, which can be expressed as:

[0065]

[0066] In the formula, represents the weighted rolling average change rate at time t; σΔE(t) represents the weighted rolling standard deviation change rate at time t; ΔE(j) represents the cooling efficiency change rate; by extracting the change rate of the cooling efficiency and its weighted rolling average and standard deviation, the dynamic change of the cooling efficiency over time is obtained; by introducing the change rate of the cooling efficiency and its weighted rolling average and standard deviation, the dynamic change of the cooling efficiency over time can be captured more accurately, which helps to identify the short-term fluctuations and long-term trends of the system performance. Taking the change rate of the cooling efficiency and its statistical characteristics as the model input can provide richer and more representative information, comprehensively considering the change characteristics of the cooling efficiency and the influence of the workload, enabling the fault prediction model to better understand the system state, so as to further improve the accuracy of the prediction result.

[0067] Step S400: Construct and train a fault prediction model based on the first statistical feature and the second statistical feature, and use the trained fault prediction model to output a fault prediction result;

[0068] Step S400 specifically includes:

[0069] Step S410: Use an LSTM network to extract time series feature representations based on the first statistical feature;

[0070] Step S410 specifically includes:

[0071] Step S411: Generate a time series dataset based on the rolling average and rolling standard deviation of the water quality index and the rolling average and rolling standard deviation of the material corrosion rate;

[0072] Step S412: At each time step, the long short-term memory network receives the input vector at the current time step and the state information at the previous time step, and updates the internal state through a gating mechanism; specifically, it receives the input vector at the current time step, the state information and the memory state at the previous time step; calculates the forgetting ratio, storage ratio and output ratio of the input vector at different times through the Sigmoid function; updates the memory state and the hidden state;

[0073] Step S413: After the LSTM processes the input vectors at all times, it outputs a vector containing the information of the entire sequence and uses this vector as the time series feature representation of the air compressor circulating water system; in this embodiment, the LSTM network can effectively capture the long-term and short-term dependencies in the time series data, thereby identifying complex fault patterns in the cooling system; converting statistical features such as rolling average and rolling standard deviation into time series feature representations can better reflect the dynamic change characteristics of the system, and the model can more accurately learn the differences between the normal operation and fault states of the system, thus improving the accuracy of fault prediction.

[0074] Step S420: Use the XGBoost model to construct a fault prediction model, train the fault prediction model based on the time series feature representation and the second statistical feature, and use the trained fault prediction model to output the fault prediction result;

[0075] Step S420 specifically includes:

[0076] Step S421: Combine the time series feature representation and the second statistical feature, which can be expressed as: The label data is defined as y t ; Initialize the XGBoost model parameters and the predicted value, and select the mean of the true labels of all samples as the initial predicted value;

[0077] Step S422: With the goal of maximizing the gain, train the fault prediction model based on the combined time series feature representation and the second statistical feature; specifically, first calculate the first-order gradient and the second-order gradient between the current predicted value and the true label, and then use the combined feature vector and the gradient to construct a decision tree. At each node, select the best split point to maximize the gain, which can be expressed as:

[0078]

[0079] In the formula, G L 、G R are the sums of the first-order gradients of the left and right child nodes respectively; H L 、H RThey are respectively the sum of the second-order gradients of the left and right child nodes; λ represents the L2 regularization term; γ represents the minimum gain required for splitting; according to the constructed decision tree, update the predicted values of each sample until the preset number of trees is reached, and add up the predicted values of all the trees to obtain the final predicted value; use the trained fault prediction model to output the fault prediction result; in this embodiment, by combining time series feature representation and statistical features such as the change rate of cooling efficiency, the XGBoost model can learn more comprehensive data patterns, thereby improving the accuracy of fault prediction; on the other hand, the XGBoost model can handle complex non-linear relationships and is suitable for scenarios such as the air compressor circulating water system with multiple variables and complex interactions; the finally output fault prediction results include but are not limited to: fault probability: predicting the probability that the air compressor circulating water system will fail in the next period of time (for example, 24 hours or 7 days); fault type: predicting that problems such as a decrease in cooling efficiency may occur in the future due to deteriorated water quality, severe scaling or corrosion; fault occurrence time: the specific time range when the fault may occur; key feature contributions and fault severity;

[0080] In summary, in this embodiment, by combining time series features and statistical features, the fault detection model can capture richer information; time series features can reflect the dynamic change trend of the data of the air compressor circulating water system, while statistical features provide the static characteristics of the data; this multi-dimensional information helps the model to more accurately understand the state of the air compressor circulating water system, ultimately improving the prediction accuracy, enhancing the robustness and generalization ability of the model.

[0081] Step S500: Based on the fault prediction results, generate a dynamic maintenance strategy to minimize downtime and maintenance costs; specifically, obtain the fault prediction results from Step S400, including fault probability, fault type, predicted fault time, key feature contribution, and fault severity; evaluate the maintenance cost and downtime cost. The maintenance cost includes labor cost, spare part cost, tool and equipment cost, etc., and different maintenance measures (such as cleaning the cooler, replacing the cooling water, replacing components, etc.) have different costs; the downtime cost includes losses caused by production interruption, costs of emergency repairs, and decreased customer satisfaction, etc. The longer the downtime, the higher the downtime cost. Then, according to the fault prediction results and historical maintenance data, formulate a preliminary maintenance plan. For example, if it is predicted that the cooling efficiency decreases and the probability is high, arrange a cooler cleaning or replace the cooling water; finally, use algorithms such as linear programming, genetic algorithm, simulated annealing, etc. to adjust the preliminary maintenance plan to achieve the goal of minimizing the total cost (maintenance cost + downtime cost); at least the following factors need to be considered during the optimization process: maintenance time window: select a suitable maintenance time window to avoid performing maintenance during peak production periods; resource allocation: reasonably allocate maintenance personnel and tools to ensure the efficient progress of maintenance work; spare part inventory: ensure that the required spare parts are in place before maintenance to reduce waiting time; multi-task coordination: if multiple maintenance tasks can be carried out simultaneously, try to arrange them to be completed within the same time period to reduce the total downtime; based on the optimized maintenance plan, generate a dynamic maintenance strategy. This strategy should detail the specific arrangements for each maintenance task, including: maintenance task list: list all the maintenance tasks that need to be executed; task priority: assign a priority to each task according to the fault severity and prediction probability; schedule: provide a detailed maintenance schedule, including the start and end times of each task; resource allocation: clarify the resources (personnel, tools, spare parts) required for each task. In this embodiment, by optimizing the maintenance plan, selecting a suitable maintenance time window, and reasonably allocating resources, the downtime of the system can be significantly reduced and production continuity can be improved;

[0082] Please refer to Figure 3It should be noted that, in order to adapt to the air compressor circulating water system fault prediction method provided by the present application, the present application has also made further improvements to the existing air compressor circulating water system, specifically, the contactor and thermal relay in the water pump main circuit of the air compressor cooling system are replaced with a frequency converter, the frequency converter can more accurately control the operating frequency of the water pump, thereby adjusting the water flow rate; the control of the frequency converter is connected to the water pump control circuit of the PLC (programmable logic controller), and the control frequency is written and started and stopped by the PLC. The specific steps are as follows: connect the control signal line of the frequency converter to the corresponding I / O port of the PLC, write the control logic in the PLC program, and dynamically adjust the output frequency of the frequency converter according to the fault prediction results or real-time monitoring data; when it is detected that there are more impurities in the water quality, reduce the output frequency of the frequency converter through the PLC, slow down the water flow rate, so as to better capture and discharge impurities; on the other hand, open 3 holes at the lowest point of the cooling circulating water pipeline of the air compressor cooling system, install a solenoid valve at each hole, and the solenoid valve controls the connection to the PLC. LC automatic control; its purpose is to write control logic in the PLC program, and automatically control the opening and closing of the solenoid valve according to the fault prediction results or real-time monitoring data; on the other hand, set a regular sewage discharge plan, or according to water quality monitoring data (such as turbidity, conductivity, etc.), automatically open the solenoid valve for sewage discharge when more impurities are detected; the air compressor circulating water system provided in this embodiment can use the XGBoost fault prediction model constructed in step S400, combined with the time series features and the second statistical features extracted by LSTM, to output the fault prediction results; based on the fault prediction results, the PLC executes the corresponding control strategy: low flow rate sewage discharge: when it is predicted that there are more impurities and the cooling efficiency may be affected, the PLC reduces the output frequency of the inverter, slows down the water flow rate, and opens the solenoid valve for sewage discharge; normal operation: when no more impurities are detected, maintain normal water flow rate and cooling efficiency, effectively reduce impurities in the system, reduce the possibility of scaling of the cooler, and thus improve the cooling efficiency.

[0083] See also Figure 2 The technical solution of the second aspect of the present invention provides a fault prediction system for a circulating water system of an air compressor, the fault prediction system comprising:

[0084] A data acquisition module, configured to obtain historical operation data of the air compressor circulating water system and perform preprocessing;

[0085] The first statistical feature extraction module is configured to extract the first statistical feature of the water quality index and the material corrosion rate in the historical operation data according to the preset time window

[0086] A second statistical feature extraction module is configured to extract a second statistical feature of cooling efficiency in combination with the workload of the air compressor

[0087] A fault prediction module, configured to construct and train a fault prediction model based on a first statistical feature and a second statistical feature, and output a fault prediction result by using the trained fault prediction model;

[0088] A policy generation module, configured to generate a dynamic maintenance policy based on the fault prediction result to minimize downtime and maintenance costs.

[0089] The technical solution of the third aspect of the present invention provides an electronic device, which includes: a processor and a memory communicatively connected to the processor; wherein, the memory stores instructions executable by the processor, and when the instructions are executed by the processor, the processor is enabled to execute the steps of the air compressor circulating water system fault prediction method according to the technical solution of the first aspect of the present invention.

[0090] The technical solution of the fourth aspect of the present invention provides a computer-readable storage medium, on which a program for implementing the air compressor circulating water system fault prediction method is stored, and when the program for implementing the air compressor circulating water system fault prediction method is executed by a processor, the steps of the air compressor circulating water system fault prediction method according to the technical solution of the first aspect of the present invention are implemented.

[0091] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only an example and does not constitute a limitation to this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are proposed in this specification, so such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments of this specification.

[0092] At the same time, this specification uses specific terms to describe the embodiments of this specification. Such as "one embodiment", "an embodiment", and / or "some embodiments" mean a certain feature, structure, or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "an embodiment" or "one embodiment" or "an alternative embodiment" mentioned twice or more at different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0093] Moreover, unless otherwise specified in the claims, the order of the processing elements and sequences, the use of numerical and alphabetical characters, or the use of other names in this specification are not intended to limit the order of the processes and methods of this specification. Although some currently useful embodiments of the invention have been discussed through various examples in the above disclosure, it should be understood that such details are for illustrative purposes only. The appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that conform to the essence and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only through software solutions, such as installing the described system on existing servers or mobile devices.

[0094] Similarly, it should be noted that, in order to simplify the presentation of the disclosure in this specification and thus assist in the understanding of one or more embodiments of the invention, in the foregoing description of the embodiments of this specification, various features are sometimes grouped together in one embodiment, drawing, or description thereof. However, this method of disclosure does not mean that the features required by the subject matter of this specification are more than those mentioned in the claims. In fact, the features of the embodiments are less than all the features of the individual embodiments disclosed above.

[0095] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification can be regarded as consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A method for predicting failures in circulating water systems of air compressors, characterized in that: The method comprises: Obtain historical operating data of the air compressor circulating water system and perform preprocessing; Extracting first statistical features of water quality indicators and material corrosion rates in historical operation data according to a preset time window; extracting a second statistical feature of cooling efficiency in combination with the workload of the air compressor; Building and training a fault prediction model based on the first statistical feature and the second statistical feature, and outputting a fault prediction result using the trained fault prediction model; Based on the failure prediction results, a dynamic maintenance strategy is generated to minimize downtime and maintenance cost.

2. The method for predicting air compressor circulating water system failure according to claim 1, characterized in that: The historical operation data of the air compressor circulating water system at least includes water quality index data, material corrosion data, inlet and outlet temperature data and workload data.

3. The method for predicting air compressor circulating water system failure according to claim 2, characterized in that: The first statistical features of water quality indicators and material corrosion rates extracted from historical operation data according to the preset time window include: Extracting water quality index vectors and material corrosion rate vectors from historical operation data according to preset time windows; The first statistical feature is extracted using the water quality index vector and the material corrosion rate vector, including: for each component in the water quality index and the material corrosion rate, respectively calculating the rolling average of the water quality index and the rolling average of the material corrosion rate; based on the rolling average of the water quality index and the rolling average of the material corrosion rate, respectively extracting the rolling standard deviation of the water quality index and the rolling standard deviation of the material corrosion rate.

4. The method for predicting air compressor circulating water system failure according to claim 2, characterized in that: The second statistical feature of cooling efficiency extracted in combination with the air compressor workload includes: Extract the cooling efficiency change rate based on the workload of the air compressor circulating water system at different times; Calculate the weighted rolling mean and standard deviation of the cooling efficiency change rate.

5. The method for predicting air compressor circulating water system failure according to any one of claims 1 to 4, characterized in that: A fault prediction model is constructed and trained based on the first statistical feature and the second statistical feature, and a fault prediction result is outputted using the trained fault prediction model, including: Using the LSTM network, the time series feature representation is extracted based on the first statistical feature; The fault prediction model is constructed using the XGBoost model, the fault prediction model is trained based on the time series feature representation and the second statistical feature, and the fault prediction result is output using the trained fault prediction model.

6. The method for predicting air compressor circulating water system failure according to claim 5, characterized in that: Using the LSTM network, the time series feature representation is extracted based on the first statistical feature, including: Generate time series data sets based on the rolling mean and rolling standard deviation of water quality indicators and the rolling mean and rolling standard deviation of material corrosion rates; At each time step, the LSTM network receives the input vector of the current time step and the state information of the previous time step, and updates the internal state through the gating mechanism; After LSTM processes the input vectors at all times, it outputs a vector containing the entire sequence information and uses the vector as the time series feature representation of the air compressor circulating water system.

7. The method for predicting air compressor circulating water system failure according to claim 5, characterized in that: The fault prediction model is constructed using the XGBoost model. The fault prediction model is trained based on the time series feature representation and the second statistical feature. The fault prediction results output by the trained fault prediction model include: Combining the time series feature representation with the second statistical feature; With the goal of maximizing gain, a fault prediction model is trained based on the combined time series feature representation and the second statistical feature; Use the trained fault prediction model to output the fault prediction results.

8. The air compressor circulating water system fault prediction system is characterized by: The fault prediction system includes: A data acquisition module, configured to obtain historical operation data of the air compressor circulating water system and perform preprocessing; The first statistical feature extraction module is configured to extract the first statistical feature of the water quality index and the material corrosion rate in the historical operation data according to the preset time window A second statistical feature extraction module is configured to extract a second statistical feature of cooling efficiency in combination with the workload of the air compressor A fault prediction module is configured to construct and train a fault prediction model based on the first statistical feature and the second statistical feature, and output a fault prediction result using the trained fault prediction model; The strategy generation module is configured to generate a dynamic maintenance strategy based on the fault prediction result to minimize downtime and maintenance cost.

9. An electronic device, characterized in that: The electronic device includes: a processor and a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and the instructions are executed by the processor so that the processor can execute the steps of the air compressor circulating water system fault prediction method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program for implementing a method for predicting a fault in a circulating water system of an air compressor, and the program for implementing a method for predicting a fault in a circulating water system of an air compressor is executed by a processor to implement the steps of the method for predicting a fault in a circulating water system of an air compressor as described in any one of claims 1 to 7.

Images