Predictive maintenance method, device and equipment for heat exchanger and readable medium

Abstract

The invention discloses a predictive maintenance method, device and equipment for a heat exchanger and a readable medium. The predictive maintenance method comprises the steps that a plurality of online operation parameters related to the equipment maintenance process of heat exchanger equipment to be maintained are obtained; carrying out preprocessing operation and data completion operation on the collected online operation parameters to obtain equipment operation parameters; building a heat exchanger steady-state heat exchange model related to heat exchange based on the equipment operation parameters, and performing flash evaporation calculation based on the heat exchanger steady-state heat exchange model to obtain various parameters of inlet and outlet material flows of the heat exchanger required for building a heat exchanger equipment mechanism model; establishing a mechanism model about the heat exchanger based on the structure information of the heat exchanger and various parameters of the inlet and outlet logistics, wherein the mechanism model of the heat exchanger is used for generating mechanism data required in the training process of the heat exchanger performance prediction model; and training the heat exchanger performance prediction model based on the mechanism data to obtain a trained heat exchanger performance prediction model, and predicting the performance of the heat exchanger based on the model.

Description

Technical Field

[0001] The present invention relates to the technical field of heat exchanger maintenance, and in particular to a predictive maintenance method, device, equipment and readable medium for a heat exchanger. Background Art

[0002] In the chemical industry, heat exchangers are key heat exchange equipment, and their performance is crucial to the entire production process. During the heat transfer process, dissolved salts or other impurities in the fluid crystallize and deposit on the surface of the heat exchanger to form scaling, which significantly reduces the heat transfer efficiency and thus reduces economic benefits. In addition, if the fluid carries sediment or solid particles, these substances will deposit inside the heat exchanger, which will not only cause the performance of the heat exchanger to deteriorate and become clogged, but in severe cases may even cause safety accidents, resulting in shutdown of the device, and further causing greater economic losses. However, due to the complex structure and invisibility of the heat exchanger itself, it is difficult to directly observe the scaling or clogging inside it during the production process. Summary of the invention

[0003] The present invention provides a predictive maintenance method, device, equipment and readable medium for a heat exchanger, so as to overcome at least one technical problem existing in the prior art.

[0004] According to a first aspect of an embodiment of the present invention, a predictive maintenance method for a heat exchanger is provided, comprising:

[0005] S1. Acquire several online operating parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the several online operating parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the several online operating parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained;

[0006] S2. Preprocessing the online operating parameters collected in step S1, including data cleaning, data labeling, and noise elimination, to ensure data quality. Meanwhile, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally the equipment operating parameters of the heat exchanger equipment after preprocessing and data completion are obtained, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment;

[0007] S3. Building a steady-state heat exchange model of the heat exchanger equipment based on the equipment operation parameters, and performing flash calculation based on the steady-state heat exchange model of the heat exchanger to obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for building a mechanism model of the heat exchanger equipment;

[0008] S4. Building a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics, wherein the mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process;

[0009] S5. The heat exchanger performance prediction model is trained based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and a plurality of performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

[0010] Preferably, the preprocessing operation in step S2 also includes data filling for missing time series data in the online operation parameters, data cleaning for abnormal data in the online operation parameters, data smoothing for the online operation parameters, and data detrending for the online operation parameters to remove the offset generated when the sensor obtains the online operation parameters.

[0011] Preferably, the heat exchanger performance prediction model uses a convolutional neural network model.

[0012] Preferably, during the training process of the convolutional neural network model, the plurality of mechanism data are divided into two parts, a training set and a validation set, according to a predetermined ratio;

[0013] The convolutional neural network is trained using the mechanism data in the training set as training samples; during the training process, the weight parameters in the convolutional neural network are updated according to the following optimizer algorithm, and the training is stopped when the accuracy of the trained convolutional neural network on the validation set meets a predetermined threshold; wherein the optimizer algorithm specifically includes:

[0014] Set the learning rate ε, decay rate ρ, momentum coefficient μ, set the initial values ​​of the parameters to be learned θ, cumulative square gradient r, cumulative gradient v in the convolutional neural network, and set the value of 10 -5 The parameter δ,

[0015] m samples are randomly selected from the training set, wherein each of the m samples carries a corresponding performance indicator, and the mechanism data of the m samples is expressed as: {x (1) ,...,x (m)}, the corresponding performance index is expressed as: {y (1) ,...,y (m)};

[0016] Calculate the average gradient g of the m samples at the current time step t according to the following rules t : Among them, the function f(x (i) ; θ) represents the function of convolutional neural network mapping, function L(f(x (i) ;θ),y (i) ) represents the loss function; Represents the gradient symbol;

[0017] Update the cumulative gradient v corresponding to the current time step t according to the following rules t :v t ←μv t-1 +(1-μ)g t , where the symbol v t-1 Represents the cumulative gradient corresponding to the previous time step of the current time step t;

[0018] Update the accumulated square gradient r corresponding to the current time step t according to the following rules t :r t ←ρr t-1 +(1-ρ)g t , where the symbol r t-1 represents the cumulative square gradient corresponding to the previous time step of the current time step t;

[0019] Calculate the first correction deviation according to the following rules and the second corrected deviation in,

[0020] Update the parameters to be learned at the current time step t based on the following rules Among them, the symbol θ t-1 represents the parameter to be learned corresponding to the previous time step of the current time step t;

[0021] Update time step t=t+1;

[0022] The training is stopped when the accuracy of the trained parameter θ on the validation set meets a predetermined threshold.

[0023] According to a second aspect of an embodiment of the present invention, there is provided a predictive maintenance device for a heat exchanger, comprising:

[0024] An online operation parameter acquisition module, used to acquire several online operation parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the several online operation parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the several online operation parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained;

[0025] A data preprocessing and data completion module is used to perform preprocessing operations including data cleaning, data labeling and noise elimination on the online operating parameters collected by the online operating parameter acquisition module to ensure the quality of the data. At the same time, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally obtain the equipment operating parameters of the heat exchanger equipment after the preprocessing and data completion operations, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment;

[0026] A parameter acquisition module for inlet and outlet logistics, used to build a steady-state heat exchange model of the heat exchanger equipment based on the equipment operation parameters, perform flash evaporation calculation based on the steady-state heat exchange model of the heat exchanger, and obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for building a mechanism model of the heat exchanger equipment;

[0027] A heat exchanger equipment mechanism model building module, used to build a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics, and the mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process;

[0028] The heat exchanger performance monitoring module is used to train the heat exchanger performance prediction model based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and several performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

[0029] According to a third aspect of an embodiment of the present invention, there is provided a predictive maintenance device for a heat exchanger, comprising:

[0030] at least one processor; and,

[0031] a memory communicatively connected to the at least one processor; wherein,

[0032] The processor stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0033] S1. Acquire several online operating parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the several online operating parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the several online operating parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained;

[0034] S2. Preprocessing the online operating parameters collected in step S1, including data cleaning, data labeling, and noise elimination, to ensure data quality. Meanwhile, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally the equipment operating parameters of the heat exchanger equipment after preprocessing and data completion are obtained, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment;

[0035] S3. Building a steady-state heat exchange model of the heat exchanger equipment based on the equipment operation parameters, and performing flash calculation based on the steady-state heat exchange model of the heat exchanger to obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for building a mechanism model of the heat exchanger equipment;

[0036] S4. Building a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics, wherein the mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process;

[0037] S5. The heat exchanger performance prediction model is trained based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and a plurality of performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

[0038] According to a fourth aspect of an embodiment of the present invention, there is provided a computer-readable medium on which computer-readable instructions are stored. The computer-readable instructions can be executed by a processor to implement the method described above.

[0039] One embodiment of this specification can achieve at least the following beneficial effects:

[0040] 1. Improved prediction accuracy: By establishing a mechanism model of the heat exchanger and combining it with actual operating data, the technical solution of the present invention can more accurately evaluate the operating status of the heat exchanger. Compared with methods that only rely on the mechanism model of the heat exchange process or pure machine learning, this combination improves the prediction accuracy, especially when facing the complexity of the production process and the errors in monitoring data.

[0041] 2. Ability to conduct early warning and maintenance: The system can continuously monitor the performance indicators of the heat exchanger in real time, such as logarithmic mean temperature difference, pressure drop, heat transfer coefficient, etc. Once performance degradation or blockage risk is found, an early warning will be issued and the predictive maintenance program will be started, thereby avoiding production interruptions caused by sudden failures, ensuring production safety and reducing economic losses.

[0042] 3. Enhanced system flexibility and adaptability: The technical solution of the present invention generates a large amount of sample data through a mechanism model to train the prediction model, reducing the dependence on the historical database, so that the system can still operate effectively in the face of insufficient data or missing historical data, thereby enhancing the system's flexibility and adaptability to new scenarios.

[0043] 4. Improved maintenance efficiency and equipment life: Through predictive maintenance, measures can be taken before the performance of the heat exchanger drops significantly, reducing unnecessary maintenance operations, while optimizing equipment performance and extending service life, thereby improving the overall economic benefits of chemical production.

[0044] 5. Enhanced interpretability: The application of mechanism models not only improves prediction accuracy, but also makes the prediction results more physically meaningful and interpretable, making it easier for technicians to understand and trust maintenance decisions, which is conducive to optimizing maintenance strategies and improving decision-making efficiency.

[0045] In summary, the technical solution of the present invention significantly improves the efficiency and effect of predictive maintenance of heat exchangers by comprehensively using mechanism models and actual operation data, providing a safer, more efficient and economical maintenance solution for chemical production. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 A schematic diagram of a process flow of a predictive maintenance method for a heat exchanger provided in an embodiment of this specification;

[0048] Figure 2 A schematic diagram of a process for predicting the performance of a heat exchanger device using the heat exchanger performance prediction model provided by the technical solution of the present invention;

[0049] Figure 3 The embodiments of this specification provide corresponding to Figure 1 A schematic diagram of a structure of a predictive maintenance device for a heat exchanger;

[0050] Figure 4 The embodiments of this specification provide corresponding to Figure 1 A schematic diagram of the structure of a predictive maintenance device for a heat exchanger. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of one or more embodiments of this specification clearer, the technical solutions of one or more embodiments of this specification will be clearly and completely described below in combination with the specific embodiments of this specification and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of one or more embodiments of this specification.

[0052] It should be understood that although the terms first, second, third, etc. may be used in this application document to describe various information, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other.

[0053] In an embodiment of the present specification, a predictive maintenance method for a heat exchanger is provided.

[0054] Next, a predictive maintenance method for a heat exchanger provided in an embodiment of the specification will be specifically described with reference to the accompanying drawings.

[0055] Figure 1 A flow chart of a predictive maintenance method for a heat exchanger provided in an embodiment of this specification is shown as follows: Figure 1 As shown, the process may include the following steps:

[0056] Step S1: Acquire several online operating parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the several online operating parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the several online operating parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained.

[0057] Generally speaking, heat exchanger equipment is a widely used equipment in chemical and other industrial fields. Its basic function is to promote heat transfer between two or more fluids, realize heat recovery or exchange, and thus meet the requirements of temperature control of different fluids in the production process. In the heat exchange process, the heat of one fluid is transferred to another fluid, and the two are not directly mixed, which ensures the stability of their respective chemical properties. The performance of heat exchanger equipment directly affects the efficiency and energy consumption of the entire production process, so its maintenance and management are particularly important.

[0058] In this embodiment, the real-time monitoring system can adopt the DCS (Distributed Control System) distributed control system widely used in industry. DCS is a computer control system based on microprocessors and microcomputers, integrating computer technology, data communication technology, CRT screen display technology and automatic control technology. In this step, historical data and real-time data can be collected from the DCS related to the heat exchanger equipment to be maintained.

[0059] Step S2: Perform preprocessing operations including data cleaning, data labeling and noise elimination on the online operating parameters collected in step S1 to ensure data quality. At the same time, if there is a necessary position in the chemical process of the heat exchanger equipment that lacks directly measured instrument data, other data associated with the instrument data that lacks directly measured are indirectly calculated to complete the instrument data that lacks directly measured, and finally obtain the equipment operating parameters of the heat exchanger equipment after preprocessing and data completion operations. The equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment.

[0060] The purpose of this step is to calculate the parameters required for chemical process simulation from the online operating parameters obtained in step S1. The role of chemical process simulation is to calculate the parameters required for heat exchanger performance prediction. This process uses the collected operating data of the real-time monitoring system (such as temperature, pressure, liquid level, flow rate, composition, etc.) to calculate through a series of physical and chemical principles to obtain the detailed thermodynamic parameters of the import and export logistics required to build the heat exchanger mechanism model, including temperature, pressure, mass enthalpy, density, viscosity, thermal conductivity and mass heat capacity, etc. These parameters are the basis for the subsequent establishment of a heat exchanger performance prediction model, which helps to more accurately evaluate the working state of the heat exchanger and predict its performance changes. In the actual scheme, the operating data of the real-time monitoring system collected in step S1 (such as temperature, pressure, liquid level, flow rate, composition, etc.) can be preprocessed first. The preprocessing work can cover operations such as data cleaning, data labeling, and noise elimination to ensure the quality of the data. In addition, if there is a lack of directly measured instrument data at a necessary location in the chemical process, it can be obtained through indirect calculation through other related data. For example, if the inlet and outlet flow rates of the heat exchanger tube are unknown, the information of the storage tank connected to the heat exchanger outlet can be used to calculate the integral of the tank level change and inversely solve the outlet flow rate of the heat exchanger. After this series of preprocessing, the obtained data will be more accurate and reliable, and suitable for parameter calculation of subsequent chemical process simulation. Through the above steps, the preprocessed and data-completed data are converted into the parameters required for chemical process simulation, which helps to understand and optimize the working performance of the heat exchanger.

[0061] Step S3: constructing a steady-state heat exchange model of the heat exchanger equipment based on the equipment operating parameters, performing flash calculation based on the steady-state heat exchange model of the heat exchanger, and obtaining various parameters of the inlet and outlet logistics of the heat exchanger equipment required for constructing the mechanism model of the heat exchanger equipment.

[0062] The role of the steady-state heat transfer model of the heat exchanger in this step is to obtain the various parameters required for building the mechanism model in the subsequent steps based on the actual operating data such as the temperature, pressure, composition, enthalpy, etc. of the inlet and outlet streams of the heat exchanger through thermodynamic calculation methods, such as flash calculation, including the temperature, pressure, mass enthalpy, density, viscosity, thermal conductivity and mass heat capacity of the inlet and outlet streams. These parameters are the basis for understanding and predicting the internal heat exchange process of the heat exchanger, and are crucial for evaluating the performance of the heat exchanger, detecting scaling or blockage, and predicting maintenance needs. Through this model, the working efficiency of the heat exchanger can be accurately analyzed without directly observing the internal state of the heat exchanger, thereby guiding the efficient operation and maintenance of the heat exchanger. When establishing a steady-state heat transfer model of the heat exchanger, it is necessary to consider the state changes of the fluid during the heat exchange process, including possible phase changes, such as liquid evaporation into steam or steam condensation into liquid. Flash calculation is the core part of processing this phase change process. It determines the amount of phase change that the fluid can undergo under given pressure and temperature conditions based on thermodynamic principles. Therefore, flash calculation is a key step in chemical process simulation, which is used to accurately calculate the phase equilibrium state of the fluid in the heat exchanger and the changes in thermodynamic parameters caused by phase change, such as enthalpy and composition. These calculation results can be used to construct and calibrate the steady-state heat transfer model of the heat exchanger to ensure that the model can accurately reflect the actual physical process inside the heat exchanger.

[0063] Specifically, flash calculation is an important part of chemical process simulation. In chemical process simulation, in order to accurately predict the operating status and performance of equipment such as heat exchangers, it is necessary to mathematically describe and calculate various physical and chemical phenomena in the process. Flash calculation specifically refers to the simulation of the process in which a fluid partially vaporizes or condenses due to changes in pressure or temperature in a two-phase or multi-phase system. It helps determine the phase equilibrium state of the fluid under specific conditions, such as the gas-liquid fraction, the composition of each phase, etc. Therefore, flash calculation is embedded in chemical process simulation to serve the overall system simulation, ensuring that the behavior of equipment such as heat exchangers under actual operating conditions can be accurately understood and predicted, such as calculating key parameters such as temperature, pressure, mass enthalpy, etc. of import and export logistics, which is crucial for establishing an accurate heat exchanger mechanism model.

[0064] Flash calculation is an important part of chemical process simulation. It is mainly used to determine the phase equilibrium state of the fluid when it experiences a sudden change in pressure, such as the process of liquid evaporating into gas or gas condensing into liquid. In the predictive maintenance system of heat exchangers based on mechanism models, flash calculation is used to obtain the detailed thermodynamic state of the heat exchanger inlet and outlet logistics. The specific steps can be taken as follows: 1. Collect necessary data: First, it is necessary to obtain basic information such as temperature, pressure, and composition of the fluid from the real-time monitoring system. These data are the basis for flash calculation. 2. Select thermodynamic method: According to the properties of the fluid and process conditions, select appropriate thermodynamic models or equations. Common ones include ideal gas model, RK equation, PR equation, etc. These models can describe the PVT (pressure-volume-temperature) relationship and phase equilibrium characteristics of the fluid. 3. Perform calculation: Using the selected thermodynamic model, calculate the phase state of the fluid (liquid phase, gas phase or two-phase coexistence) under given pressure and temperature conditions. For two-phase fluids, the calculation will determine the respective mole fraction, mass fraction, phase fraction, and thermodynamic properties such as enthalpy, entropy, and density of each phase. 4. Application software tools: In actual operation, these calculations are usually completed with the help of professional chemical process simulation software such as Aspen Plus, HYSYS, PRO / II, etc. These softwares have built-in multiple thermodynamic packages and flash calculation modules. Users only need to input the initial conditions, and the software can automatically perform flash calculations and output the results. 5. Application of results: The results of flash calculations will be directly used for further construction of heat exchanger mechanism models and performance predictions, such as determining key parameters such as temperature, pressure, mass enthalpy, density, viscosity, thermal conductivity and mass heat capacity of import and export logistics, which is crucial for accurately evaluating the operating status of the heat exchanger. In summary, flash calculation is a calculation method that predicts the phase change process of fluids under specific conditions through thermodynamic models and professional software. It is an indispensable part of building heat exchanger mechanism models and performance predictions.

[0065] Step S4: Building a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics. The mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process.

[0066] In this step, the mechanism model is an accurate mathematical model established based on the internal mechanism of the object, production process or the transfer mechanism of material flow. This type of model is obtained based on basic principles such as mass balance equation, energy balance equation, momentum balance equation, phase balance equation, physical property equation, chemical reaction law, etc. The parameters have clear physical meanings and can accurately describe the behavior and characteristics of the system. In the predictive maintenance system of heat exchangers, the mechanism model is used to simulate the heat exchange process of the heat exchanger and its performance changes, which is the basis for building a predictive model. The equipment mechanism model is based on the steady-state heat exchange model of the heat exchanger, and further incorporates the specific design data and structural information of the heat exchanger equipment, such as the type, material, size, and fluid characteristics of the shell and tube side of the heat exchanger, to accurately simulate the actual working performance of a specific heat exchanger equipment. In short, the equipment mechanism model is a refinement and concretization of the steady-state heat exchange model of the heat exchanger, which can provide more accurate performance predictions for actual equipment.

[0067] The mechanism model simulates the internal working principle and heat exchange process of the heat exchanger, and uses the design data, real-time monitoring data and thermodynamic principles of the heat exchanger to calculate the performance of the heat exchanger under various working conditions. Specifically, first, according to the structural information of the heat exchanger (such as the TEMA type of the heat exchanger, detailed information of the baffle: material, cut direction, baffle spacing, etc., shell information: installation angle, material, thickness, corrosion allowance, geometric data, etc., tube information: type, material, geometric data, number of tubes, distribution, etc., nozzle information: installation position, geometric data, flow direction, etc.) and the temperature, pressure, composition and other parameters of the inlet and outlet streams, the necessary logistics characteristics (such as temperature, pressure, enthalpy, etc.) are obtained with the help of thermodynamic calculation methods such as flash calculation. Then, these data are used to perform chemical process simulation in the selected commercial process simulation software to obtain more detailed heat exchanger operating parameters. The large amount of data generated by this series of calculations reflects the expected performance of the heat exchanger under different conditions, which constitutes the so-called "mechanism data". These data are subsequently used to train models such as artificial neural networks to achieve predictive maintenance of heat exchanger performance. In practice, commercial software such as Aspen Exchanger Design and Rating, HTRI, etc. can be used to establish a heat exchanger equipment mechanism model, which will take into account the complex flow and heat transfer mechanism inside the heat exchanger and provide a basis for subsequent predictions.

[0068] The mechanism model obtained in this step can autonomously generate a large amount of sample data. Even in the face of insufficient data or loss of historical data, the heat exchanger performance prediction model obtained in the subsequent steps can complete the performance prediction task without over-reliance on the historical database, which significantly improves the flexibility of the system and makes it more interpretable.

[0069] Step S5: The heat exchanger performance prediction model is trained based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and a plurality of performance indicators of the heat exchanger equipment are calculated, and when the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

[0070] In this step, a large amount of mechanism data generated by chemical process simulation is used to fit the heat exchanger performance prediction model. For example, an artificial neural network model can be used to adjust the model's hyperparameters (such as the number of hidden layers, batch size, learning rate, etc.) so that the model can accurately predict the performance indicators of the heat exchanger, such as logarithmic mean temperature difference, pressure drop, heat transfer coefficient, etc.

[0071] Through the above steps, a mechanism model that reflects the internal working principle of the heat exchanger can be established to provide a scientific basis for predictive maintenance. This step realizes the continuous real-time evaluation of the equipment operation status and the accurate prediction of the future potential state. Before the heat exchanger fails and affects production, it can help process engineers and equipment engineers determine the cause of the failure, intervene in advance in time, confirm whether the intervention means are effective, and ensure the continuous operation of the production process.

[0072] The steady-state heat transfer model of the heat exchanger mentioned in the previous article focuses on describing the heat exchange process of the heat exchanger under steady-state conditions. It is based on the principles of thermodynamics and calculates the heat transfer between the hot fluid and the cold fluid inside the heat exchanger through parameters such as temperature, pressure, and enthalpy. This model is mainly used to understand the basic physical laws of the heat exchange process, but does not involve the specific structural details of the heat exchanger. The mechanism model of the heat exchanger equipment goes a step further. In addition to considering the basic principles of heat exchange, it also incorporates the detailed mechanism of the flow characteristics inside the heat exchanger, such as the resistance, stratification, turbulence and other effects of fluid flow, as well as the convection and radiation heat transfer mechanisms on the heat exchange surface. This model can more accurately reflect the actual working performance of the heat exchanger and provide a theoretical basis for prediction and optimization. The mechanism model of the heat exchanger equipment is designed for a specific heat exchanger and contains all the physical structure information of the heat exchanger, such as the type of heat exchanger (such as TEMA type), baffle design, shell and tube material and size, etc. It uses these specific design parameters, combined with the heat exchange mechanism model, to conduct a more accurate performance evaluation. The equipment mechanism model can reflect the performance changes caused by differences in the heat exchanger structure, and is the basis for personalized maintenance and optimization. In short, the steady-state heat exchange model of the heat exchanger focuses on the basic principles of heat exchange. The mechanism model of the heat exchanger equipment adds flow and heat exchange details on this basis, and further integrates the structural characteristics of the specific equipment. The three are progressive and serve together for the performance prediction and maintenance of the heat exchanger.

[0073] Figure 1The technical effects of the technical solution in the invention are mainly reflected in the following aspects: 1. Improved prediction accuracy: By establishing a mechanism model of the heat exchanger and combining it with actual operation data, the technical solution of the invention can more accurately evaluate the operating status of the heat exchanger. Compared with the method that only relies on the mechanism model of the heat exchange process or pure machine learning, this combination improves the accuracy of prediction, especially when facing the complexity of the production process and the error of monitoring data. 2. Ability to provide early warning and maintenance: The system can continuously monitor the performance indicators of the heat exchanger in real time, such as logarithmic mean temperature difference, pressure drop, heat transfer coefficient, etc. Once performance degradation or blockage risk is found, it will issue an early warning and start the predictive maintenance program, thereby avoiding production interruptions caused by sudden failures, ensuring production safety and reducing economic losses. 3. Enhanced system flexibility and adaptability: The technical solution of the invention generates a large amount of sample data through the mechanism model to train the prediction model, reducing the dependence on the historical database, so that the system can still operate effectively in the face of insufficient data or missing historical data, enhancing the flexibility of the system and its adaptability to new scenarios. 4. Improved maintenance efficiency and equipment life: Through predictive maintenance, measures can be taken before the performance of the heat exchanger drops significantly, reducing unnecessary maintenance operations, while optimizing the performance of the equipment and extending its service life, improving the overall economic benefits of chemical production. 5. Enhanced interpretability: The application of the mechanism model not only improves the accuracy of the prediction, but also makes the prediction results have stronger physical meaning and interpretability, making it easier for technicians to understand and trust maintenance decisions, which is conducive to optimizing maintenance strategies and improving decision-making efficiency.

[0074] In summary, the technical solution of the present invention significantly improves the efficiency and effect of predictive maintenance of heat exchangers by comprehensively using mechanism models and actual operation data, providing a safer, more efficient and economical maintenance solution for chemical production.

[0075] based on Figure 1 The method, the examples of this specification also provide some specific implementation plans of the method, which are described below.

[0076] In an optional embodiment technical solution, the preprocessing operation in step S2 may also include data filling for missing time series data in the online operation parameters, data cleaning for abnormal data in the online operation parameters, data smoothing for the online operation parameters, and data detrending for the online operation parameters to remove the offset generated when the sensor acquires the online operation parameters.

[0077] Meanwhile, the heat exchanger performance prediction model may use a convolutional neural network model. Accordingly, in an optional embodiment technical solution, during the training process of the convolutional neural network model, the plurality of mechanism data are divided into a training set and a validation set according to a predetermined ratio;

[0078] The convolutional neural network is trained using the mechanism data in the training set as training samples; during the training process, the weight parameters in the convolutional neural network are updated according to the following optimizer algorithm, and the training is stopped when the accuracy of the trained convolutional neural network on the validation set meets a predetermined threshold; wherein the optimizer algorithm specifically includes:

[0079] Set the learning rate ε, decay rate ρ, momentum coefficient μ, set the initial values ​​of the parameters to be learned θ, cumulative square gradient r, cumulative gradient v in the convolutional neural network, and set the value of 10 -5 The parameter δ,

[0080] m samples are randomly selected from the training set, wherein each of the m samples carries a corresponding performance indicator, and the mechanism data of the m samples is expressed as: {x (1) ,...,x (m)}, the corresponding performance index is expressed as: {y (1) ,...,y (m)};

[0081] Calculate the average gradient g of the m samples at the current time step t according to the following rules t : Among them, the function f(x (i) ; θ) represents the function of convolutional neural network mapping, function L(f(x (i) ;θ),y (i) ) represents the loss function; Represents the gradient symbol;

[0082] Update the cumulative gradient v corresponding to the current time step t according to the following rules t :v t ←μv t-1 +(1-μ)g t , where the symbol v t-1 Represents the cumulative gradient corresponding to the previous time step of the current time step t;

[0083] Update the accumulated square gradient r corresponding to the current time step t according to the following rules t :r t ←ρr t-1 +(1-ρ)g t , where the symbol r t-1 represents the cumulative square gradient corresponding to the previous time step of the current time step t;

[0084] Calculate the first correction deviation according to the following rules and the second corrected deviation in,

[0085] Update the parameters to be learned at the current time step t based on the following rules Among them, the symbol θ t-1 represents the parameter to be learned corresponding to the previous time step of the current time step t;

[0086] Update time step t=t+1;

[0087] The training is stopped when the accuracy of the trained parameter θ on the validation set meets a predetermined threshold.

[0088] Based on the same idea, the embodiments of this specification also provide a device corresponding to the above method. Figure 3 The embodiments of this specification provide corresponding to Figure 1 A schematic diagram of the structure of a predictive maintenance device for a heat exchanger. Figure 3 As shown, the device may include:

[0089] The online operation parameter acquisition module 302 is used to obtain several online operation parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, and the several online operation parameters are obtained from the real-time monitoring system corresponding to the heat exchanger equipment to be maintained. The several online operation parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of the substances in the heat exchanger equipment to be maintained.

[0090] The data preprocessing and data completion module 304 is used to perform preprocessing operations including data cleaning, data labeling and noise elimination on the online operating parameters collected by the online operating parameter acquisition module to ensure the quality of the data. At the same time, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally the equipment operating parameters of the heat exchanger equipment after the preprocessing and data completion operations are obtained, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment for the heat exchanger equipment.

[0091] The parameter acquisition module 306 for inlet and outlet logistics is used to build a steady-state heat exchange model of the heat exchanger equipment based on the equipment operating parameters, perform flash evaporation calculations based on the steady-state heat exchange model of the heat exchanger, and obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for building a mechanism model of the heat exchanger equipment.

[0092] The heat exchanger equipment mechanism model building module 308 is used to build a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics. The mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process.

[0093] The heat exchanger performance monitoring module 310 is used to train the heat exchanger performance prediction model based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and a plurality of performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

[0094] It is understood that the above modules refer to computer programs or program segments for executing one or more specific functions. In addition, the distinction between the above modules does not mean that the actual program codes must also be separated.

[0095] Based on the same idea, the embodiments of this specification also provide a device corresponding to the above method. Figure 4 The embodiments of this specification provide corresponding to Figure 1 A schematic diagram of the structure of a predictive maintenance device for a heat exchanger. Figure 4 As shown, the device 400 may include:

[0096] at least one processor 410; and,

[0097] A memory 430 is communicatively connected to the at least one processor; wherein,

[0098] The memory 430 stores instructions 420 that can be executed by the at least one processor 410 , and the instructions are executed by the at least one processor 410 .

[0099] The instructions may enable the at least one processor 410 to:

[0100] Acquire a number of online operating parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the number of online operating parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the number of online operating parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained;

[0101] The online operating parameters collected in the previous step are preprocessed, including data cleaning, data labeling, and noise elimination, to ensure data quality. Meanwhile, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally the equipment operating parameters of the heat exchanger equipment after the preprocessing and data completion operations are obtained, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment;

[0102] A steady-state heat exchange model of the heat exchanger equipment is constructed based on the equipment operation parameters, and flash evaporation calculation is performed based on the steady-state heat exchange model of the heat exchanger to obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for constructing the mechanism model of the heat exchanger equipment;

[0103] Building a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics, wherein the mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process;

[0104] The heat exchanger performance prediction model is trained based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and a plurality of performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

[0105] Based on the same idea, the embodiments of this specification also provide a computer-readable medium corresponding to the above method. The computer-readable medium stores computer-readable instructions, which can be executed by a processor to implement the predictive maintenance method of the heat exchanger provided in the above embodiments.

[0106] Those skilled in the art can understand that the accompanying drawings are only schematic diagrams of an embodiment, and the modules or processes in the accompanying drawings are not necessarily required to implement the present invention.

[0107] Those skilled in the art can understand that the modules in the device in the embodiment can be distributed in the device in the embodiment according to the description of the embodiment, or can be changed accordingly and located in one or more devices different from the embodiment. The modules in the above embodiment can be combined into one module, or can be further divided into multiple sub-modules.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A predictive maintenance method for a heat exchanger, characterized in that: The method comprises: S1. Acquire several online operating parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the several online operating parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the several online operating parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained; S2. Preprocessing the online operating parameters collected in step S1, including data cleaning, data labeling, and noise elimination, to ensure data quality. Meanwhile, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally the equipment operating parameters of the heat exchanger equipment after preprocessing and data completion are obtained, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment; S3. Building a steady-state heat exchange model of the heat exchanger equipment based on the equipment operation parameters, and performing flash calculation based on the steady-state heat exchange model of the heat exchanger to obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for building a mechanism model of the heat exchanger equipment; S4. Building a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics, wherein the mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process; S5. The heat exchanger performance prediction model is trained based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and a plurality of performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

2. The predictive maintenance method for a heat exchanger according to claim 1, characterized in that: The preprocessing operation in step S2 also includes data filling for missing time series data in the online operation parameters, data cleaning for abnormal data in the online operation parameters, data smoothing for the online operation parameters, and data detrending for the online operation parameters to remove the offset generated when the sensor obtains the online operation parameters.

3. The predictive maintenance method for a heat exchanger according to claim 1, characterized in that: The heat exchanger performance prediction model uses a convolutional neural network model.

4. The predictive maintenance method for a heat exchanger according to claim 3, characterized in that: During the training process of the convolutional neural network model, the plurality of mechanism data are divided into a training set and a validation set according to a predetermined ratio; The convolutional neural network is trained using the mechanism data in the training set as training samples; during the training process, the weight parameters in the convolutional neural network are updated according to the following optimizer algorithm, and the training is stopped when the accuracy of the trained convolutional neural network on the validation set meets a predetermined threshold; wherein the optimizer algorithm specifically includes: Set the learning rate ε, decay rate ρ, momentum coefficient μ, set the initial values ​​of the parameters to be learned θ, cumulative square gradient r, cumulative gradient v in the convolutional neural network, and set the value of 10 -5 The parameter δ, m samples are randomly selected from the training set, wherein each of the m samples carries a corresponding performance indicator, and the mechanism data of the m samples is expressed as: {x (1) ,...,x (m) }, the corresponding performance index is expressed as: {y (1) ,...,y (m) }; Calculate the average gradient g of the m samples at the current time step t according to the following rules t : Among them, the function f(x (i) ; θ) represents the function of convolutional neural network mapping, function L(f(x (i) ;θ),y (i) ) represents the loss function; Represents the gradient symbol; Update the cumulative gradient v corresponding to the current time step t according to the following rules t :v t ←μv t-1 +(1-μ)g t , where the symbol v t-1 Represents the cumulative gradient corresponding to the previous time step of the current time step t; Update the accumulated square gradient r corresponding to the current time step t according to the following rules t :r t ←ρr t-1 +(1-ρ)g t , where the symbol r t-1 represents the cumulative square gradient corresponding to the previous time step of the current time step t; Calculate the first correction deviation according to the following rules and the second corrected deviation in, Update the parameters to be learned at the current time step t based on the following rules Among them, the symbol θ t-1 represents the parameter to be learned corresponding to the previous time step of the current time step t; Update time step t=t+1; Repeat the iteration until the parameters to be learned are trained. θ The training is stopped when the accuracy on the validation set meets a predetermined threshold.

5. A predictive maintenance device for a heat exchanger, characterized in that: include: An online operation parameter acquisition module, used to acquire several online operation parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the several online operation parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the several online operation parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained; A data preprocessing and data completion module is used to perform preprocessing operations including data cleaning, data labeling and noise elimination on the online operating parameters collected by the online operating parameter acquisition module to ensure the quality of the data. At the same time, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally obtain the equipment operating parameters of the heat exchanger equipment after the preprocessing and data completion operations, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment; A parameter acquisition module for inlet and outlet logistics, used to build a steady-state heat exchange model of the heat exchanger equipment based on the equipment operation parameters, perform flash evaporation calculation based on the steady-state heat exchange model of the heat exchanger, and obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for building a mechanism model of the heat exchanger equipment; A heat exchanger equipment mechanism model building module, used to build a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics, and the mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process; The heat exchanger performance monitoring module is used to train the heat exchanger performance prediction model based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and several performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

6. A predictive maintenance device for a heat exchanger, comprising: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The processor stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire a number of online operating parameters related to the equipment maintenance process of the heat exchanger equipment to be maintained, wherein the number of online operating parameters are acquired from a real-time monitoring system corresponding to the heat exchanger equipment to be maintained, and the number of online operating parameters include temperature parameters, pressure parameters, liquid level parameters, flow parameters of the heat exchanger equipment to be maintained, and composition parameters of substances in the heat exchanger equipment to be maintained; The online operating parameters collected in the previous step are preprocessed, including data cleaning, data labeling, and noise elimination, to ensure data quality. Meanwhile, if there is a lack of directly measured instrument data at a necessary position in the chemical process of the heat exchanger equipment, other data associated with the instrument data lacking direct measurement are indirectly calculated to complete the instrument data lacking direct measurement, and finally the equipment operating parameters of the heat exchanger equipment after the preprocessing and data completion operations are obtained, and the equipment operating parameters are used to build a steady-state heat exchange model of the heat exchanger equipment; A steady-state heat exchange model of the heat exchanger equipment is constructed based on the equipment operation parameters, and flash evaporation calculation is performed based on the steady-state heat exchange model of the heat exchanger to obtain various parameters of the inlet and outlet logistics of the heat exchanger equipment required for constructing the mechanism model of the heat exchanger equipment; Building a mechanism model of the heat exchanger equipment based on the structural information of the heat exchanger equipment and various parameters of the import and export logistics, wherein the mechanism model of the heat exchanger equipment is used to generate a number of mechanism data required by the heat exchanger performance prediction model during the training process; The heat exchanger performance prediction model is trained based on the plurality of mechanism data to obtain a trained heat exchanger performance prediction model; then, basic measurement point information is collected from the real-time monitoring system at preset time intervals, the basic measurement point information is brought into the heat exchanger performance prediction model, and a plurality of performance indicators of the heat exchanger equipment are calculated. When the value of at least one of the plurality of performance indicators is monitored to exceed a predetermined threshold range corresponding to any one of the performance indicators, corresponding maintenance processing is performed on the heat exchanger equipment.

7. A computer-readable medium having computer-readable instructions stored thereon, wherein the computer-readable instructions can be executed by a processor to implement the predictive maintenance method for a heat exchanger according to any one of claims 1 to 4.

Images