Fuel cell air compressor heat dissipation method, device, and equipment

Abstract

The present invention provides a method, device, and equipment for heat dissipation of a fuel cell air compressor. The method comprises: hierarchically collecting a multidimensional thermal data set of a fuel cell air compressor; performing feature extraction on the multidimensional thermal data set, inputting the extracted feature data set into a neural network model based on deep transfer learning, and obtaining a corresponding heat dissipation efficiency attenuation prediction model; establishing a multi-objective optimization equation for the air compressor based on the multidimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solving the multi-objective optimization equation to obtain a heat dissipation control strategy set; dynamically partitioning the heat dissipation system of the air compressor, and collaboratively adjusting the flow rate and flow direction of the coolant in each dynamically partitioned area according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor. The present invention solves the problems of heat dissipation efficiency attenuation and uneven resource allocation caused by changes in coolant quality and integrated control in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment heat dissipation, and in particular to a heat dissipation method, device and equipment for a fuel cell air compressor. Background Art

[0002] Fuel cells are highly efficient and clean energy conversion devices that are widely used in new energy vehicles and distributed power generation systems. As a core auxiliary device in fuel cell systems, air compressors are primarily responsible for providing the appropriate amount of compressed air to the fuel cell stack. Their operational stability directly impacts the overall performance of the fuel cell.

[0003] During the operation of a fuel cell air compressor, a large amount of heat is generated due to high-speed rotation and compression, requiring heat management through a cooling system. Existing methods for cooling fuel cell air compressors primarily employ a single coolant circulation system, managing heat through a fixed cooling strategy and unified cooling control parameters. This cooling method presents the following technical issues: Due to the lack of real-time monitoring and consideration of changes in coolant quality, cooling efficiency decreases significantly when coolant quality changes. Furthermore, the traditional cooling system employs an integrated control strategy that is unable to precisely manage the differences in heat loads within different areas of the air compressor, resulting in uneven distribution of cooling resources, which not only affects the cooling effect but also wastes energy. Summary of the Invention

[0004] The main purpose of the present invention is to solve the technical problems of heat dissipation efficiency degradation and uneven resource allocation caused by the integrated control strategy used in existing fuel cell air compressor heat dissipation systems, which cannot precisely manage the heat load differences in different areas within the air compressor.

[0005] A first aspect of the present invention provides a fuel cell air compressor heat dissipation method, the fuel cell air compressor heat dissipation method comprising:

[0006] The direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor are collected in layers to obtain a multidimensional thermal data set of the air compressor;

[0007] Extracting features from the multidimensional thermal data set, and inputting the extracted feature data set into a neural network model based on deep transfer learning to obtain a corresponding heat dissipation efficiency attenuation prediction model;

[0008] Establishing a multi-objective optimization equation for the air compressor based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solving the multi-objective optimization equation to obtain a heat dissipation control strategy set;

[0009] The heat dissipation system of the air compressor is dynamically partitioned, and the flow rate and flow direction of the coolant in each area of the dynamic partition are coordinated and adjusted according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor.

[0010] Optionally, in a first implementation of the first aspect of the present invention, the layered collection of direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor to obtain a multidimensional thermal data set of the air compressor includes:

[0011] Setting a first layer of data collection points at a heat exchange position where the coolant of the fuel cell air compressor flows through, to collect the direct thermal characteristic parameters, wherein the direct thermal characteristic parameters include temperature data, flow rate data, and pressure data;

[0012] Setting a second layer of data collection points at the preset core components of the air compressor to collect the indirect thermal characteristic parameters, wherein the indirect thermal characteristic parameters include speed data, current data, and vibration data;

[0013] A third layer of data collection points is set in the coolant circulation system of the air compressor's heat dissipation system to collect coolant quality characteristic parameters, including conductivity, pH value and turbidity data;

[0014] Time synchronization and data preprocessing are performed on the direct thermal characteristic parameters, the indirect thermal characteristic parameters, and the coolant quality characteristic parameters to obtain a multidimensional thermal data set of the air compressor.

[0015] Optionally, in a second implementation of the first aspect of the present invention, extracting features from the multidimensional thermal dataset, inputting the extracted feature dataset into a neural network model based on deep transfer learning, and obtaining a corresponding heat dissipation efficiency attenuation prediction model includes:

[0016] Performing time domain feature extraction, frequency domain feature extraction, and time-frequency feature extraction on each parameter in the multidimensional thermal data set to obtain target domain features;

[0017] The heat dissipation efficiency attenuation pattern features of the air compressor in the preset source domain knowledge base are used as source domain features, and the feature mapping algorithm of the attention mechanism is used to perform adaptive conversion between the source domain features and the target domain features;

[0018] Fusing the target domain features with the adapted source domain features, and dynamically assigning weights to the fused features through a hierarchical attention network to obtain a feature dataset;

[0019] The feature data set is input into a neural network model based on deep transfer learning for training, and the heat dissipation efficiency attenuation prediction model is obtained by minimizing the distribution difference between the source domain features and the target domain features and the heat dissipation efficiency prediction error.

[0020] Optionally, in a third implementation of the first aspect of the present invention, establishing a multi-objective optimization equation for the air compressor based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solving the multi-objective optimization equation to obtain a heat dissipation control strategy set includes:

[0021] Establishing a thermodynamic model of the air compressor heat dissipation system according to the multidimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and constructing the multi-objective optimization equation based on the thermodynamic model;

[0022] Establishing a set of constraints for the multi-objective optimization equation, wherein the set of constraints includes upper temperature limits for each component, coolant flow range constraints, pump power limit constraints, and dynamic constraints based on coolant quality characteristic parameters;

[0023] Inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer to solve and obtain a Pareto optimal solution set;

[0024] The current operating condition is determined according to the multidimensional thermal data set, and the Pareto optimal solution set is dynamically screened according to the current operating condition, and a corresponding heat dissipation control strategy set is generated according to the screened Pareto optimal solution set.

[0025] Optionally, in a fourth implementation of the first aspect of the present invention, inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer to solve and obtain a Pareto optimal solution set includes:

[0026] Inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer, and mapping the objective function space of the multi-objective optimization equation into the decision variable space by the multi-objective dung beetle optimizer to obtain an initial solution population of the optimization problem;

[0027] Calculating the objective function value for each solution in the initial solution population to obtain an evaluation value matrix of the solution, calculating the dominance relationship between any two solutions based on the evaluation value matrix, hierarchically sorting the solutions based on the dominance relationship, and obtaining a position update vector for each solution;

[0028] Applying the position update vector to each solution in the initial solution population to generate an updated solution set, and performing constraint checking on the solutions in the updated solution set. When a solution in the updated solution set violates the set of constraints, correcting the position of the corresponding solution using a preset penalty function to obtain a candidate solution set that satisfies the constraints;

[0029] Calculating the crowding distance of the candidate solution set, sorting the solutions with the same non-dominated level based on the crowding distance, and selecting high-quality solutions through the elite retention strategy to obtain the non-dominated solution set in the iterative process;

[0030] According to the preset iterative termination condition, the optimization process of the multi-objective dung beetle optimizer is cyclically iterated. When the termination condition is met, the Pareto front is extracted from the final non-dominated solution set to obtain the Pareto optimal solution set.

[0031] Optionally, in a fifth implementation of the first aspect of the present invention, dynamically partitioning the heat dissipation system of the air compressor and collaboratively adjusting the flow rate and flow direction of the coolant in each area of the dynamic partition according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor includes:

[0032] extracting temperature field distribution information according to direct thermal characteristic parameters in the multidimensional thermal data set, and calculating heat flow characteristics according to the direct thermal characteristic parameters and the indirect thermal characteristic parameters;

[0033] Clustering the temperature field distribution information and heat flow characteristics using an adaptive K-means clustering algorithm, and dynamically partitioning the heat dissipation system of the air compressor based on the clustering results to obtain N control sub-areas, where N is an integer from 2 to 8;

[0034] Calculating a heat load index for each of the control sub-regions based on the direct thermal characteristic parameters, and constructing an inter-region heat transfer matrix to characterize the heat transfer relationship between the control sub-regions;

[0035] Determining the cooling resource allocation strategy for each of the control sub-areas using a hierarchical progressive control method according to the heat dissipation control strategy set, the heat load index, and the heat transfer matrix;

[0036] The coolant flow regulating valve and flow direction control valve in the heat dissipation system are used to adjust the coolant parameters of each control sub-area in real time according to the cooling resource allocation strategy, and the cooling intensity of adjacent areas is collaboratively controlled based on the heat transfer matrix to achieve heat dissipation of the air compressor.

[0037] Optionally, in a sixth implementation of the first aspect of the present invention, adjusting the coolant parameters of each of the control sub-areas in real time according to the cooling resource allocation strategy through the coolant flow regulating valve and the flow direction control valve in the heat dissipation system, and collaboratively controlling the cooling intensity of adjacent areas based on the heat transfer matrix includes:

[0038] Prioritizing each of the control sub-areas according to the heat load index, allocating a control cycle and a response order according to the ranking result, and obtaining a control priority list;

[0039] Performing proportional-integral-differential control calculation on the flow parameters and flow direction parameters in the cooling resource allocation strategy, converting the control signal into a valve opening instruction, and obtaining a control sequence of the actuator;

[0040] Calculating the thermal coupling influence coefficient between each of the control sub-regions based on the heat transfer matrix, and adjusting the cooling resource allocation of adjacent regions according to the thermal coupling influence coefficient to obtain a collaborative control compensation amount;

[0041] The control sequence and the coordinated control compensation are integrated to generate valve control instructions, and real-time closed-loop correction is performed according to the dynamic response of the system to obtain control signals for the coolant flow control valve and the flow direction control valve;

[0042] The control signal is applied to the coolant flow regulating valve and the flow direction control valve to adjust the coolant parameters of each of the control sub-areas in real time.

[0043] A second aspect of the present invention provides a heat dissipation device for a fuel cell air compressor, the heat dissipation device for a fuel cell air compressor comprising:

[0044] A data acquisition module is used to perform layered acquisition of direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor to obtain a multidimensional thermal data set of the air compressor;

[0045] A pattern recognition module is used to extract features from the multidimensional thermal data set, input the extracted feature data set into a neural network model based on deep transfer learning, and obtain a corresponding heat dissipation efficiency attenuation prediction model;

[0046] a strategy generation module, configured to establish a multi-objective optimization equation for the air compressor based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solve the multi-objective optimization equation to obtain a heat dissipation control strategy set;

[0047] The partition control module is used to dynamically partition the heat dissipation system of the air compressor and coordinately adjust the flow rate and flow direction of the coolant in each area of the dynamic partition according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor.

[0048] The third aspect of the present invention provides a fuel cell air compressor heat dissipation device, comprising: a memory and at least one processor, wherein the memory stores instructions, and the memory and the at least one processor are interconnected through lines; the at least one processor calls the instructions in the memory so that the fuel cell air compressor heat dissipation device performs the steps of the above-mentioned fuel cell air compressor heat dissipation method.

[0049] The above-mentioned fuel cell air compressor heat dissipation method, device, and equipment collect the multidimensional thermal data set of the fuel cell air compressor in layers; perform feature extraction on the multidimensional thermal data set, input the extracted feature data set into a neural network model based on deep transfer learning, and obtain a corresponding heat dissipation efficiency attenuation prediction model; establish a multi-objective optimization equation for the air compressor based on the multidimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solve the multi-objective optimization equation to obtain a heat dissipation control strategy set; dynamically partition the heat dissipation system of the air compressor, and coordinately adjust the flow rate and flow direction of the coolant in each area of the dynamic partition according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor. The present invention solves the problems of heat dissipation efficiency attenuation and uneven resource allocation caused by changes in coolant quality and integrated control in the prior art.

[0050] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or understood by practicing the present invention. The purposes and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0051] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a schematic diagram of a first embodiment of a method for cooling a fuel cell air compressor according to an embodiment of the present invention;

[0053] Figure 2 A schematic diagram of an embodiment of a heat dissipation device for a fuel cell air compressor according to an embodiment of the present invention;

[0054] Figure 3 This is a schematic diagram of an embodiment of a heat dissipation device for a fuel cell air compressor in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0056] The terms "including," "having," and any variations thereof, as used in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to the process, method, product, or device.

[0057] To facilitate understanding of this embodiment, a fuel cell air compressor heat dissipation method disclosed in an embodiment of the present invention is first introduced in detail. Figure 1 As shown, this method includes the following steps:

[0058] 101. Perform layered collection of direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor to obtain a multidimensional thermal data set of the air compressor;

[0059] In one embodiment of the present invention, the layered collection of direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor to obtain a multidimensional thermal data set of the air compressor includes: setting a first layer of data collection points at a heat exchange position through which the coolant of the fuel cell air compressor flows to collect the direct thermal characteristic parameters, wherein the direct thermal characteristic parameters include temperature data, flow rate data, and pressure data; setting a second layer of data collection points at a preset core component of the air compressor to collect the indirect thermal characteristic parameters, wherein the indirect thermal characteristic parameters include speed data, current data, and vibration data;

[0060] A third layer of data collection points is set in the coolant circulation system of the air compressor's heat dissipation system to collect coolant quality characteristic parameters, which include conductivity, pH value and turbidity data. The direct thermal characteristic parameters, the indirect thermal characteristic parameters and the coolant quality characteristic parameters are time synchronized and data preprocessed to obtain a multidimensional thermal dataset of the air compressor.

[0061] Specifically, the process for collecting multi-dimensional thermal signature data for the air compressor begins by establishing a first layer of data collection points at key heat exchange locations where the fuel cell air compressor coolant flows. These key heat exchange locations include the coolant inlet and outlet, the radiator interface, and areas of concentrated heat load. High-precision temperature sensors, flowmeters, and pressure sensors are installed at these locations to collect direct thermal signature parameters. Temperature data is collected using a PT100 platinum resistance temperature sensor with a measurement range of -50°C to 150°C and an accuracy of ±0.1°C. The sampling frequency is set to 100Hz, enabling the capture of subtle temperature fluctuations. Flow velocity data is measured using an ultrasonic flowmeter with a measurement range of 0.5m / s to 10m / s and an accuracy of ±1%. The sampling frequency is also set at 100Hz to ensure real-time monitoring of fluid dynamics. Pressure data is collected using a piezoresistive pressure sensor with a measurement range of 0MPa to 2MPa, an accuracy of ±0.5%, and a sampling frequency of 50Hz. These three direct thermal signature parameters constitute the fundamental physical quantities of the heat exchange process and directly reflect the heat exchange efficiency and operating status of the cooling system.

[0062] Specifically, a second layer of data collection points is set up around the compressor's core components. These include the motor's stator, rotor, bearings, and key structures such as the air inlet and compression chamber. Hall sensors are installed on the motor's stator and rotor to collect motor speed data, with a measurement range of 0 to 20,000 rpm, an accuracy of ±0.5%, and a sampling frequency of 50 Hz. Current transformers are installed in the motor's power supply lines to measure current data, with a range of 0 to 100 A, an accuracy of ±0.5%, and a sampling frequency of 50 Hz. Piezoelectric accelerometers are installed on the compressor's bearings and housing to collect vibration data, with a measurement range of 0 Hz to 10 kHz, a sensitivity of 100 mV / g, and a sampling frequency of 100 Hz to capture high-frequency vibration characteristics. While these indirect thermal signature parameters do not directly measure heat, they are closely related to the compressor's load, efficiency, and heat generation. Using these parameters, the causes and patterns of heat generation can be deduced.

[0063] Specifically, a third layer of data collection points was established within the coolant circulation system of the air compressor's cooling system, primarily located at the coolant storage tank, the circulation pump inlet, and the main loop. Conductivity sensors, pH detectors, and turbidity meters were installed at these locations to collect characteristic coolant quality parameters. The conductivity sensor uses a four-electrode method to measure the coolant's conductivity, with a measurement range of 0 μS / cm to 2000 μS / cm, an accuracy of ±1%, and a sampling frequency of 1 Hz. The pH detector uses a glass electrode method to measure the coolant's pH, with a measurement range of 0 to 14, an accuracy of ±0.1, and a sampling frequency of 1 Hz. The turbidity meter uses a scattered light measurement principle, with a measurement range of 0 to 100 NTU, an accuracy of ±2%, and a sampling frequency of 1 Hz. The frequency of collecting coolant quality parameters is low because these parameters typically change slowly and do not require frequent monitoring. These parameters reflect the coolant's contamination, corrosiveness, and thermal conductivity, which directly impact the efficiency and lifespan of the cooling system.

[0064] Specifically, time synchronization and data preprocessing of all collected data are key steps in constructing a multidimensional thermal data set. Time synchronization uses a unified timestamp marking method to align all sensor data according to the system's unified clock, eliminating deviations in sampling times between different sensors. Data preprocessing includes three steps: outlier detection, missing value filling, and data normalization. Outlier detection uses the triple standard deviation rule, marking data outside the range of ±3σ from the mean as outliers and correcting them with a sliding window median replacement method. Missing value filling uses a combination of linear interpolation and forward filling to supplement missing points in the time series. Data normalization uses the minimum-maximum standardization method to uniformly map feature data of different dimensions and orders of magnitude to the [0, 1] interval, facilitating subsequent feature extraction and model training.

[0065] 102. Extract features from the multidimensional thermal data set, input the extracted feature data set into a neural network model based on deep transfer learning, and obtain a corresponding heat dissipation efficiency attenuation prediction model;

[0066] In one embodiment of the present invention, the feature extraction of the multidimensional thermal data set and the inputting of the extracted feature data set into a neural network model based on deep transfer learning to obtain a corresponding heat dissipation efficiency attenuation prediction model include: performing time domain feature extraction, frequency domain feature extraction and time-frequency feature extraction on each parameter in the multidimensional thermal data set to obtain target domain features; using the heat dissipation efficiency attenuation pattern features of the existing air compressor in the preset source domain knowledge base as source domain features, and adopting the feature mapping algorithm of the attention mechanism to perform adaptive conversion on the source domain features and the target domain features; fusing the target domain features with the adaptively converted source domain features, and dynamically assigning weights to the fused features through a hierarchical attention network to obtain a feature data set; inputting the feature data set into a neural network model based on deep transfer learning for training, and obtaining the heat dissipation efficiency attenuation prediction model by minimizing the distribution difference and heat dissipation efficiency prediction error between the source domain features and the target domain features.

[0067] Specifically, feature extraction for multidimensional thermal datasets begins with time-domain feature extraction, frequency-domain feature extraction, and time-frequency feature extraction. During time-domain feature extraction, the system calculates statistical features for direct thermal feature parameters, indirect thermal feature parameters, and coolant quality feature parameters, including mean, standard deviation, skewness, kurtosis, interquartile range, and entropy. Taking temperature data as an example, the system uses every 10 seconds of data as a time window and calculates the mean temperature within the window to reflect the overall temperature level, the standard deviation to reflect the degree of temperature fluctuation, and the skewness and kurtosis to reflect the asymmetry and peak level of the temperature distribution, respectively. For signals with large fluctuations, such as current and vibration, the crest factor, form factor, and pulse factor are additionally calculated to capture the transient characteristics of the signal. For parameters such as flow rate and pressure, the rate of change and trend characteristics are extracted, characterized by calculating first-order differences and linear regression coefficients. Although the coolant quality parameters are sampled at a lower frequency, a sliding window method is used to extract 24-hour trend characteristics. These time domain characteristics together constitute a comprehensive description of the thermal characteristics of the air compressor in the time dimension, which can effectively reflect the basic statistical characteristics and change laws of each parameter.

[0068] Specifically, frequency domain feature extraction focuses on analyzing the periodic variation patterns of each parameter. The system performs fast Fourier transform (FFT) on temperature, flow rate, pressure, speed, current, and vibration data, converting the time domain signals into the frequency domain. The frequency domain features extracted for each parameter include the main frequency and its amplitude, power spectrum density, frequency band energy distribution, and spectral moment. Specifically, the system decomposes the vibration signal into three frequency bands: 0-100Hz, 100-1000Hz, and 1000-10000Hz, respectively, extracting low-frequency, medium-frequency, and high-frequency vibration energy. These features correspond to the vibration characteristics of different components of the air compressor. Temperature and flow rate data focus on the spectral features in the 0-1Hz range, extracting the frequency values and amplitudes of the first five main frequency components, as well as the energy proportion in the 0.1-0.5Hz range. Frequency domain analysis of current data focuses on the power supply frequency (such as 50Hz) and its harmonic components, reflecting the working status of the motor by calculating the total harmonic distortion rate. These frequency domain features can identify periodic patterns hidden in time domain data and play an important role in indicating the attenuation of heat dissipation efficiency caused by bearing failures, motor anomalies, etc.

[0069] Specifically, time-frequency feature extraction utilizes a combination of wavelet transform and Hilbert-Huang transform to capture the non-stationary characteristics of parameters at different time scales. The system performs wavelet decomposition on signals such as temperature, flow rate, and vibration, using the Daubechies (db4) wavelet as the basis function. The signals are decomposed into five scales, and the wavelet energy entropy, wavelet packet entropy, and singular value decomposition features are extracted for each scale. For vibration signals, empirical mode decomposition (EMD) is additionally applied to decompose the signals into several intrinsic mode functions (IMFs), extracting the instantaneous frequency and instantaneous amplitude features of each IMF. For temperature and flow rate data, continuous wavelet transform is used to calculate time-frequency spectrograms, from which time-frequency energy distribution features are extracted. These time-frequency features can capture changes in the energy distribution of the signal at different time scales and are significantly advantageous for detecting sudden and gradual changes in the operating state of the air compressor. Through these three feature extraction methods, the system extracts hundreds of dimensions of target domain features from a multidimensional thermal dataset, comprehensively characterizing the thermal characteristics of the current air compressor.

[0070] Specifically, the core step in transfer learning is to adapt and transform the existing heat dissipation efficiency decay pattern characteristics of air compressors in a pre-set source domain knowledge base using an attention mechanism feature mapping algorithm. The source domain knowledge base stores historical operating data of multiple air compressors under different operating conditions and different usage durations, along with their corresponding heat dissipation efficiency decay characteristics. These characteristics also include time domain, frequency domain, and time-frequency features. The attention mechanism feature mapping algorithm first aligns the source and target domain features in feature space and quantifies the distribution difference between the two domains using the maximum mean difference (MMD) criterion. The algorithm then constructs a multi-head attention network, with each attention head responsible for learning the importance of a specific type of feature, such as focusing on temperature-related features and vibration-related features. Each attention head calculates the similarity matrix between the query vector (from the target domain) and the key vector (from the source domain), generates attention weights, and then performs a weighted summation on the value vector (source domain features) based on this weighted weight to obtain the adapted feature representation. The outputs of the multi-head attention are concatenated and linearly transformed to obtain the final source domain feature adaptation result. This attention-based feature mapping method can adaptively learn the feature correspondence between the source domain and the target domain, deal with the distribution offset problem between the two domains, and achieve effective knowledge transfer.

[0071] Specifically, fusing target domain features with adapted source domain features requires a carefully designed hierarchical attention network for dynamic weight assignment. This fusion process first groups features according to their physical meaning, such as temperature, fluid, mechanical, and electrical groups. Each group contains relevant time, frequency, and time-frequency features. The hierarchical attention network employs a three-layer architecture: the bottom layer focuses on intra-group feature fusion, the middle layer focuses on inter-group feature fusion, and the top layer focuses on the fusion of source and target domain features. At the bottom layer, a self-attention mechanism is employed within each feature group to learn the relationships between different representations of the same physical quantity (e.g., time, frequency, and time-frequency). At the middle layer, a cross-attention mechanism is used to learn the mutual influence between different feature groups, such as the impact of temperature changes on current changes. At the top layer, a gated fusion unit dynamically adjusts the weights of source and target domain features. In its implementation, each layer's attention computation includes three steps: attention score calculation, softmax normalization, and weighted summation. The fused features are then further enhanced through residual connections and layer normalization to enhance their expressive power. This hierarchical attention fusion method fully considers the physical correlation and hierarchical structure between features, so that the fused feature dataset not only retains the specificity of the target domain, but also utilizes the rich knowledge of the source domain, providing high-quality input for subsequent model training.

[0072] Specifically, inputting the feature dataset into a neural network model based on deep transfer learning for training requires a carefully designed network structure and training strategy. This neural network model adopts an encoder-decoder architecture. The encoder consists of three main modules: a feature extraction module, a domain adaptation module, and a task-specific module. The feature extraction module consists of four stacked bidirectional GRUs (Gated Recurrent Units), with 128 neurons per layer, capturing the long-term and short-term dependencies of temporal features. The domain adaptation module uses a combination of gradient reversal layers and a domain classifier, using adversarial training to ensure that the extracted features are invariant to the source and target domains. The task-specific module consists of two fully connected layers, with 256 and 128 neurons per layer, respectively, using the GELU activation function. The decoder adopts a multi-task learning framework to simultaneously predict the heat dissipation efficiency value and the decay trend category. During training, the loss function consists of four components: a mean squared error loss for heat dissipation efficiency prediction, a cross-entropy loss for decay trend classification, a maximum mean difference loss for feature distributions, and an adversarial loss for domain classification. Training uses the Adam optimizer with a batch size of 64, an initial learning rate of 0.001, and a cosine annealing strategy for dynamic learning rate adjustment. To prevent overfitting, weight decay and early stopping mechanisms are introduced. During training, the weight of source domain data gradually decreases with increasing iterations, while the weight of target domain data gradually increases, achieving a smooth transition from source domain knowledge-based to target domain adaptation-based. This deep transfer learning approach enables the training of an accurate heat dissipation efficiency degradation prediction model, even with limited target domain data. This model not only predicts the specific heat dissipation efficiency value but also determines the trend and severity of degradation.

[0073] 103. Based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, a multi-objective optimization equation for the air compressor is established, and the heat dissipation control strategy set is obtained by solving the multi-objective optimization equation;

[0074] In one embodiment of the present invention, the multi-objective optimization equation of the air compressor is established based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and the heat dissipation control strategy set is obtained by solving the multi-objective optimization equation, including: establishing a thermodynamic model of the air compressor heat dissipation system according to the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and constructing the multi-objective optimization equation based on the thermodynamic model; establishing a constraint condition set of the multi-objective optimization equation, wherein the constraint condition set includes upper temperature limit constraints of each component, coolant flow range constraints, pump power limit constraints and dynamic constraints based on coolant quality characteristic parameters; inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer for solving to obtain a Pareto optimal solution set; determining the current operating conditions according to the multi-dimensional thermal data set, and dynamically screening the Pareto optimal solution set according to the current operating conditions, and generating a corresponding heat dissipation control strategy set according to the screened Pareto optimal solution set.

[0075] Specifically, the strategy generation process requires the development of a thermodynamic model for the air compressor cooling system, leveraging the outputs of a multidimensional thermal dataset and a heat dissipation efficiency decay prediction model. This thermodynamic model comprises three submodules: a heat conduction module, a heat convection module, and a heat radiation module. The heat conduction module describes heat transfer between solid components within the compressor. Based on Fourier heat conduction theory, it simplifies the compressor into multiple thermal particles, each representing a key component such as the motor stator, rotor, or bearing. The thermal resistance between these particles is calculated based on the component material thermal conductivity, contact area, and distance, while also accounting for the nonlinear effect of component temperature on the thermal resistance. The heat convection module describes the heat exchange between solid components and the coolant, modeled using Newton's law of cooling. The convective heat transfer coefficient of the coolant flowing through each heat exchange surface is determined by dimensionless parameters such as the Reynolds number and the Prandtl number. For areas with large variations in flow velocity, the system uses computational fluid dynamics to pre-simulate the convective heat transfer coefficient at different flow rates and creates a lookup table for real-time calculations. The thermal radiation module describes the radiative heat transfer between components. Although this factor accounts for a relatively small proportion in water-cooled systems, it still needs to be considered in high-temperature areas and is calculated using the Stefan-Boltzmann law. The parameters of these three submodules are optimized using historical data from a multidimensional thermal dataset. The parameters are dynamically adjusted based on efficiency trends predicted by a heat dissipation efficiency decay prediction model to reflect the system's time-varying characteristics. The resulting thermodynamic model accurately describes the temperature distribution and heat flow of various air compressor components under different operating conditions.

[0076] Specifically, based on the established thermodynamic model, a multi-objective optimization equation was constructed. This optimization equation contains three conflicting objectives: maximizing heat dissipation efficiency, minimizing energy consumption, and maximizing equipment life. The heat dissipation efficiency maximization function defines the degree to which the average temperature of key components deviates from the target temperature. The closer the temperature is to the target value, the higher the heat dissipation efficiency. The target temperature varies depending on the component type, such as 75°C for the motor stator and 65°C for the bearing. The energy consumption minimization function calculates the total energy consumption of the cooling system, including the power consumption of the coolant pump and the control system. Pump power consumption is proportional to flow rate and pressure and is calculated using the pump characteristic curve. The equipment life maximization function is constructed based on the Arrhenius equation, influencing the temperature fluctuation amplitude, temperature gradient, and number of temperature cycles of key components, and modeling their relationship with the material aging rate. It is often difficult to achieve optimality for all three objectives simultaneously. For example, improving heat dissipation efficiency typically requires increasing coolant flow, which increases energy consumption. Reducing temperature fluctuations can extend equipment life, but may require more frequent adjustments and increase the burden on the control system. The decision variables of the optimization equation include the coolant flow rate in each area, the opening of the flow control valve, and the operation frequency of the heat dissipation control. These variables together determine the operating strategy of the heat dissipation system.

[0077] Specifically, solving the multi-objective optimization equations requires a series of constraints to ensure the practical feasibility of the solution. The system's established constraint set first includes upper temperature limits for each component, such as motor stator temperature not exceeding 120°C and bearing temperature not exceeding 80°C. These constraints are hard constraints based on material heat resistance and safety standards and must not be violated under any circumstances. Coolant flow range constraints specify the upper and lower flow limits for each cooling circuit, such as 10-50 L / min for the main circuit and 5-20 L / min for the branch circuit, taking into account pipe diameter, pump capacity, and fluid dynamics. Pump power limit constraints ensure that the required pump power does not exceed the equipment rating, preventing overload. An innovative feature is the dynamic constraint based on coolant quality parameters, which adjusts the maximum allowable flow rate in real time based on the coolant's conductivity, pH, and turbidity. For example, when the turbidity exceeds 30 NTU, the maximum flow rate in the main circuit is reduced from 50 L / min to 30 L / min to reduce contaminant deposition and pipe corrosion under high-speed flow. When the pH value is below 6.5 or above 8.5, the permissible range of temperature fluctuations is reduced to reduce the risk of corrosion. This set of constraints ensures the practical operability of the optimization results and system safety, while also considering the potential impact of changes in coolant quality on the cooling system.

[0078] Specifically, inputting the multi-objective optimization equation and its set of constraints into a pre-defined multi-objective dung beetle optimizer for solution is a key step in obtaining a Pareto-optimal solution set. The multi-objective dung beetle optimizer is a swarm intelligence algorithm inspired by the natural behavior of dung beetles in finding, rolling, and burying dung balls. It is particularly well-suited for solving complex optimization problems with multiple conflicting objectives. The optimizer's workflow first encodes the decision variables of the optimization problem as position vectors of individual dung beetles. The initial population consists of 200 randomly distributed individuals. In each iteration, the algorithm simulates three dung beetle behaviors: foraging, ball rolling, and burying. Foraging corresponds to global search, in which each dung beetle updates its search direction based on its own position and the global optimal position. The search radius decreases with each iteration, achieving a smooth transition from exploration to exploitation. Ball rolling corresponds to local refinement, in which the dung beetle rolls the solution it has found along the gradient, deciding whether to accept the new position by comparing the non-dominated relationship between the new and old positions. Burial behavior corresponds to the preservation and diversity of solutions. The algorithm maintains an external archive of non-dominated solutions and maintains the diversity of the solution set through meshing and crowding distance calculation. When processing constraints, an improved constraint processing technique is used to convert the degree of constraint violation into a penalty term and add it to the objective function. The optimization process sets a maximum number of iterations to 1000, and the convergence condition is that the change in the Pareto front is less than a preset threshold after 50 consecutive iterations. Ultimately, the optimizer outputs a set of evenly distributed Pareto optimal solutions, each representing a parameter configuration for a heat dissipation control strategy. These solutions perform differently on the three objective functions, making it impossible to simply declare one as the absolute optimal.

[0079] Specifically, after obtaining the Pareto optimal solution set, it is necessary to select the most appropriate solution as the actual control strategy based on the current operating conditions. The current operating condition is determined by extracting data from a recent window (typically 10 minutes) from a multidimensional thermal dataset, calculating the mean, standard deviation, and trend of each key parameter, and constructing an operating condition feature vector. The operating condition feature vector contains information such as the motor load status, ambient temperature conditions, and coolant quality status, comprehensively reflecting the current system operating environment. A pre-trained random forest classifier is used to identify and classify the operating condition, classifying the current operating condition into predefined operating condition categories such as low-load steady state, high-load steady state, load fluctuation state, temperature increase state, and temperature decrease state. Different operating condition categories correspond to different cooling demand priorities. For example, in a high-load steady state, cooling efficiency has the highest priority; in a fluctuating state with frequent starts and stops, equipment life has the highest priority; and in low ambient temperatures, energy consumption has a higher priority. Based on the identified operating condition categories, the system calculates the weight coefficients of the three objective functions to form a comprehensive evaluation function. Each solution in the Pareto optimal solution set is then scored using this evaluation function, and the top N solutions with the highest scores (typically N = 3-5) are selected to form a candidate solution set. Finally, the system considers the requirement for smooth control strategy switching to avoid drastic changes in control parameters. By comparing the control strategy with the previous one, the solution with the smallest change and meeting the performance requirements is selected as the final cooling control strategy. This strategy includes parameters such as the coolant flow setpoint for each zone, the flow control valve opening, and the control period, forming a complete cooling control strategy set.

[0080] Furthermore, the multi-objective optimization equation and the constraint condition set thereof are input into a preset multi-objective dung beetle optimizer for solving to obtain a Pareto optimal solution set, including: inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer, mapping the objective function space of the multi-objective optimization equation to the decision variable space through the multi-objective dung beetle optimizer, and obtaining an initial solution population of the optimization problem; calculating the objective function value for each solution in the initial solution population to obtain an evaluation value matrix of the solution, and calculating the dominance relationship between any two solutions based on the evaluation value matrix, hierarchically sorting the solutions based on the dominance relationship, and obtaining a position update vector for each solution; and updating the position A vector is applied to each solution in the initial solution population to generate an updated solution set, and a constraint check is performed on the solutions in the updated solution set. When a solution in the updated solution set violates the constraint condition set, the position of the corresponding solution is corrected by a preset penalty function to obtain a candidate solution set that satisfies the constraints; a crowding distance is calculated for the candidate solution set, solutions with the same non-dominated level are sorted based on the crowding distance, and high-quality solutions are selected by an elite retention strategy to obtain a non-dominated solution set in an iterative process; according to a preset iterative termination condition, the optimization process of the multi-objective dung beetle optimizer is cyclically iterated. When the termination condition is met, the Pareto front is extracted from the final non-dominated solution set to obtain the Pareto optimal solution set.

[0081] Specifically, in this embodiment, inputting the multi-objective optimization equation and its set of constraints into a pre-set multi-objective dung beetle optimizer is the starting point for solving complex heat dissipation control problems. The multi-objective dung beetle optimizer first receives three objective functions: maximizing heat dissipation efficiency, minimizing energy consumption, and maximizing equipment lifespan, as well as a complete set of constraints, including upper temperature limits, flow range constraints, power limits, and dynamic coolant quality constraints. The optimizer uses an encoding mechanism to transform this high-dimensional optimization problem into a form suitable for processing by the dung beetle algorithm. The mapping from the objective function space to the decision variable space utilizes a direct encoding approach, representing each decision variable (such as the coolant flow rate in each region, the flow control valve opening, etc.) as a dimension in the position vector of an individual dung beetle. Continuous variables, such as flow rates, are directly encoded using real numbers; discrete variables, such as valve opening positions, are encoded using integers. This encoding approach preserves the physical meaning of the decision variables, facilitating the subsequent interpretation and application of the solution. The initial solution population is generated using Latin hypercube sampling, generating 200 candidate solutions uniformly distributed across the decision space. This sampling method more effectively covers the decision space than pure random sampling, increasing the diversity of the initial population. Furthermore, a small amount of Gaussian noise is added to each initial solution to prevent solutions from being identical in certain dimensions, further enhancing diversity. A reflection strategy is employed to handle boundaries. When a sample point exceeds the permissible range of a decision variable, it is reflected back along the boundary into the feasible region, ensuring that all initial solutions fall within the legal range of the decision variable.

[0082] Specifically, calculating the objective function value for each solution in the initial population is a fundamental step in evaluating solution quality. The system inputs the decision variable configuration represented by each solution into the previously constructed thermodynamic model, and simulates and calculates the corresponding heat dissipation efficiency, energy consumption, and equipment life indicators. Heat dissipation efficiency is calculated by the mean absolute error of the temperature deviation of key components from the target temperature; a smaller value indicates higher heat dissipation efficiency. Energy consumption is calculated by comprehensively calculating factors such as pump power and control system power consumption. Equipment life estimates the material aging rate based on factors such as temperature cycles and temperature gradients. These three indicators form the objective function evaluation vector for each solution. The evaluation vectors of all solutions together form an evaluation value matrix with a dimension of [number of solutions × number of objective functions]. Based on the evaluation value matrix, the algorithm calculates the dominance relationship between any two solutions. In multi-objective optimization, if solution A is non-inferior to solution B in all objective functions (i.e., its objective function values are not worse than solution B) and is strictly superior to solution B in at least one objective function, then solution A is said to dominate solution B. For example, if solution A's heat dissipation efficiency and energy consumption are both non-inferior to solution B, and its equipment life indicator is strictly superior to solution B, then solution A dominates solution B. By traversing and comparing all pairs of solutions, the number of times each solution is dominated (the domination count) and the set of solutions it dominates (the dominating set) are calculated. Based on the domination counts, all solutions are sorted into a non-dominated order. Solutions with a domination count of 0 (i.e., solutions not dominated by any other solution) are assigned to the first frontier (rank 1). Then, solutions with a domination count of 0 are searched for from the remaining solutions and assigned to the second frontier (rank 2). This process continues until all solutions have been assigned a rank. Based on this ranking, combined with the position update mechanism of the dung beetle optimization algorithm, the position update vector of each solution is calculated. The calculation of the position update vector consists of three steps: a global optimal guidance term, a random exploration term, and a local refinement term. The global optimal guidance term is derived from a randomly selected solution in the current non-dominated solution set. The random exploration term incorporates appropriate random perturbations. The local refinement term considers learning the optimal position of the solution in the past.

[0083] Specifically, applying the position update vector to each solution in the initial solution population is the core iterative step of the dung beetle optimization algorithm. For each solution, a new candidate position is generated based on its current position and the calculated position update vector. The position update adopts an update strategy with inertia weight, that is, the new position is not only affected by the current update vector, but also partially retains the information of the original position. This strategy helps to balance the exploration and development capabilities of the algorithm. The inertia weight is dynamically adjusted with the number of iterations. The initial value is set to 0.9 and linearly decreases to 0.4 as the iteration proceeds, prompting the algorithm to gradually shift from global exploration to local refinement. The updated solution set needs to undergo strict constraint checking to ensure that all constraints are met. When a solution violates a constraint, the solution is corrected using a static penalty method. Specifically, for solutions that violate the upper temperature limit constraint, the amount exceeding the upper temperature limit is calculated and the coolant flow rate is increased proportionally until the temperature constraint is satisfied. For solutions that violate the flow range constraint, the flow rate value that exceeds the range is directly truncated to the boundary value. For solutions that violate the pump power limit, the flow rate in each region is proportionally reduced until the total power meets the limit. For solutions that violate the dynamic coolant quality constraint, the flow rate upper limit is dynamically adjusted according to the current coolant quality parameters, and the excess is truncated. This constraint handling method not only ensures the feasibility of the solution but also improves the convergence efficiency of the algorithm through a correction strategy with clear physical meaning. After strict constraint checking, every solution in the candidate solution set satisfies all constraints, ensuring the practical operability of the optimization results.

[0084] Specifically, calculating the crowding distance of candidate solutions is an important means of maintaining solution diversity. The crowding distance reflects the density of solutions in the target space; a larger distance indicates a sparser solution population. The calculation process first sorts solutions by each objective function value and then calculates the normalized distance between neighboring solutions on that objective function. The final crowding distance of a solution is the sum of the distances across all objective functions. Boundary solutions (i.e., solutions that achieve an extreme value on any objective function) are assigned an infinite crowding distance to ensure their retention. Based on the crowding distance and the previously described non-dominated rank, the system implements an elitist retention strategy. This strategy first merges the parent and child solution sets. Solutions are then selected sequentially based on their non-dominated rank, from lowest to highest (lower ranks are preferred), until the upper limit of the population size is reached. If selecting a solution at a certain rank would result in exceeding the population size, solutions within that rank are selected from the highest to the lowest crowding distance, prioritizing solutions in sparsely distributed areas. This elitist retention strategy ensures the retention of high-quality solutions and a uniform distribution of the solution set, preventing the algorithm from prematurely converging to a local Pareto front. In practical implementation, a fast non-dominated sorting algorithm is used to improve sorting efficiency, which is particularly important for large-scale problems. The solutions selected by the elite retention strategy constitute the next generation of non-dominated solution sets. These solutions achieve different balances on the three objective functions and represent different heat dissipation control strategy choices.

[0085] Specifically, the iterative process of the multi-objective dung beetle optimizer loops according to preset termination criteria. These termination criteria are based on three criteria: a maximum number of iterations (1000), Pareto front stability, and a computational time limit. Pareto front stability is measured by a change rate of less than 0.1% in the Pareto front's hypervolume over 50 consecutive iterations. This metric comprehensively reflects the convergence and diversity of the non-dominated solution set. In each iteration, the system records the current non-dominated solution set and updates the global Pareto front. Convergence during the iterative process is characterized by a rapid movement of the Pareto front toward the ideal point of the objective function in the early stages, a continuous expansion of the front boundary in the middle stages, and a gradual uniformity of the front distribution in the later stages. To improve algorithm stability, an adaptive control parameter strategy is employed to dynamically adjust key parameters of the dung beetle optimizer, such as the rolling ball radius and burial depth, based on the search progress. When any of the termination criteria are met, the algorithm stops iterations and extracts the Pareto front from the final non-dominated solution set. This front consists of solutions that are not dominated by any other solution and represents the optimal set of trade-offs achievable under the current problem setting. In practice, to facilitate the selection and implementation of control strategies, the extracted Pareto front is post-processed, including solution sparsification (selecting representative solutions through clustering) and solution mapping (mapping solutions in the decision space back to specific control parameter configurations). The resulting Pareto optimal solution set typically contains 30-50 different solutions, each representing a specific cooling control strategy that achieves varying degrees of balance between cooling efficiency, energy consumption, and equipment lifespan.

[0086] 104. Dynamically partition the heat dissipation system of the air compressor, and coordinately adjust the flow rate and flow direction of the coolant in each area of the dynamic partition according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor.

[0087] In one embodiment of the present invention, the heat dissipation system of the air compressor is dynamically partitioned, and the flow rate and flow direction of the coolant in each area of the dynamic partition are coordinated to adjust according to the heat dissipation control strategy set to achieve the heat dissipation of the air compressor, including: extracting temperature field distribution information according to the direct thermal characteristic parameters in the multidimensional thermal data set, and calculating the heat flow characteristics according to the direct thermal characteristic parameters and the indirect thermal characteristic parameters; clustering the temperature field distribution information and heat flow characteristics using an adaptive K-means clustering algorithm, and dynamically partitioning the heat dissipation system of the air compressor based on the clustering results to obtain N control sub-areas, where N is an integer from 2 to 8. ; Based on the direct thermal characteristic parameters, the heat load index is calculated for each of the control sub-areas, and an inter-area heat transfer matrix is constructed to characterize the heat transfer relationship between the control sub-areas; according to the heat dissipation control strategy set, the heat load index and the heat transfer matrix, a hierarchical progressive control method is used to determine the cooling resource allocation strategy for each of the control sub-areas; the coolant flow regulating valve and the flow direction control valve in the heat dissipation system are used to adjust the coolant parameters of each of the control sub-areas in real time according to the cooling resource allocation strategy, and the cooling intensity of adjacent areas is coordinated controlled based on the heat transfer matrix to achieve heat dissipation of the air compressor.

[0088] Specifically, dynamic partitioning of the air compressor's cooling system first requires extracting temperature field distribution information from a multidimensional thermal dataset. This step is achieved by processing the temperature data from direct thermal signature parameters, involving data denoising, interpolation, and reconstruction. Data denoising utilizes a wavelet transform, using the db4 wavelet basis function to perform a five-layer decomposition of the temperature signal, removing high-frequency noise and reconstructing the signal to produce smooth temperature data. Interpolation utilizes radial basis functions (RBFs) to expand discrete temperature sampling points into a continuous temperature field. The interpolation grid density is set to 5 mm x 5 mm, sufficient to capture temperature gradients within the compressor. This process provides complete temperature field distribution information within the compressor, including surface and internal temperatures and temperature gradients of each component. Simultaneously, the system also calculates heat flow characteristics, utilizing both direct and indirect thermal signature parameters. Heat flow characteristic calculation utilizes a heat flux estimation method, combining direct thermal signature parameters such as flow rate and pressure with indirect thermal signature parameters such as speed and current. The heat flux distribution of each compressor component is then inverted using a thermodynamic model. Specifically, the heat flux of the motor is calculated based on current, resistance, and motor efficiency; the heat flux of the compression chamber is calculated based on compression ratio, intake temperature, and air flow; and the heat flux of the bearing is estimated based on speed, load, and friction coefficient. These heat flow characteristics describe the generation and flow of heat in the air compressor and, together with the temperature field distribution information, form the data foundation for dynamic partitioning.

[0089] Specifically, clustering temperature field distribution information and heat flow characteristics using an adaptive K-means clustering algorithm is the core technology for achieving dynamic partitioning. Traditional K-means algorithms have problems such as difficulty in determining the number of clusters K, sensitivity to initial cluster centers, and unsuitability for non-spherical distributions. Therefore, this embodiment adopts an improved adaptive K-means clustering algorithm. The algorithm first normalizes the temperature field and heat flow characteristic data to eliminate the dimensional differences between different physical quantities. Then, the optimal number of clusters K is automatically determined using the silhouette coefficient method. The specific process is to try K values from 2 to 10, perform clustering and calculate the silhouette coefficient respectively, and select the K value with the largest silhouette coefficient as the final number of clusters. To address the problem of sensitivity of the initial cluster center, the algorithm uses the density peak method to select the initial center point, that is, select a point with high local density and a long distance from the high-density point as the initial cluster center. During the iterative process, the system dynamically adjusts the feature weights, assigning higher weights to areas with large temperature gradients, so that hot spots can form independent clusters. The clustering process ends when the clustering results are stable or the maximum number of iterations (usually set to 100) is reached. The final clustering results divide the compressor's physical space into N control subregions, with N typically ranging from 2 to 8, depending on the compressor's complexity and operating conditions. Each control subregion has similar temperature and heat flow characteristics, making it suitable for a unified cooling control strategy. The partitioning results are dynamically adjusted as the compressor's operating conditions change. For example, under high load conditions, the motor area may be divided into multiple subregions to achieve more refined control.

[0090] Specifically, calculating the heat load index for each control sub-region based on direct thermal characteristic parameters is a key step in assessing the cooling needs of each region. The heat load index is a comprehensive indicator that reflects the urgency and importance of cooling a region. The calculation process first extracts the temperature characteristics of each control sub-region, including the average temperature, maximum temperature, temperature gradient, and temperature change rate. The average temperature is calculated by taking the arithmetic average of all temperature sampling points within the region. The maximum temperature is directly taken as the maximum temperature within the region. The temperature gradient is calculated using the finite difference method as the modulus of the spatial temperature derivative. The temperature change rate is calculated by dividing the temperature change within the time window by the time interval. The system then assigns temperature weight coefficients based on the component's temperature resistance rating, with higher weights assigned to critical components such as motor stators and bearings. The final heat load index is calculated using a weighted summation method; higher values indicate more urgent cooling needs for the region. Simultaneously, the system constructs an inter-region heat transfer matrix to describe the heat exchange relationships between control sub-regions. This matrix is an N×N square matrix, with each element representing the heat transfer coefficient from region i to region j. The heat transfer coefficient is calculated based on the contact area, material thermal conductivity, distance between the two regions, and convective heat transfer conditions. For example, the heat transfer coefficient between two adjacent metal parts is high, while the heat transfer coefficient between areas separated by insulating materials is low. This heat transfer matrix reflects the flow path and intensity of heat within the air compressor, providing an important basis for subsequent coordinated control.

[0091] Specifically, determining the cooling resource allocation strategy for each control sub-area based on the cooling control strategy set, heat load index, and heat transfer matrix is a key step in achieving precise cooling. This step utilizes a hierarchical progressive control approach, dividing the control process into three levels: the macro-overall strategy layer, the meso-regional coordination layer, and the micro-execution control layer. At the macro-overall strategy layer, the system selects the optimal strategy for the current operating conditions from the cooling control strategy set and determines the allocation principle and control target for total cooling resources (such as total flow and total power). At the meso-regional coordination layer, the system performs preliminary resource allocation based on the heat load index of each control sub-area, with regions with higher heat load indexes receiving more cooling resources. This preliminary allocation is then modified using the heat transfer matrix to account for the impact of heat exchange between regions. For example, if region A transfers a large amount of heat to region B, the cooling resources allocated to region A will be appropriately increased, or the cooling sequence between the two regions will be adjusted. This modification process uses an iterative calculation method until the predicted temperatures of each region meet the control target. At the micro-execution control layer, the system converts the resource allocation results into specific control parameters, including the flow setpoint, valve opening, and flow direction switching timing for each region. The generation of control parameters takes into account the physical limitations and dynamic characteristics of the actuators, such as valve response time and flowmeter measurement errors. This approach improves control accuracy through a combined feedforward and feedback approach. The resulting cooling resource allocation strategy not only meets overall heat dissipation requirements but also accommodates the individual needs of each area, achieving optimal cooling resource allocation.

[0092] Specifically, the execution phase of cooling control involves real-time adjustment of coolant parameters for each controlled sub-zone via the coolant flow control valves and flow direction control valves in the cooling system based on the cooling resource allocation strategy. The system first smoothes the cooling resource allocation strategy to prevent sudden changes in control parameters that could cause system oscillation. This smoothing process uses a combination of sliding average and gradient limiting to ensure the continuity and stability of the control signal. The processed control signals are transmitted via a communication interface to the various actuators, including the variable frequency pump, electric control valve, and electromagnetic directional valve. The variable frequency pump adjusts its speed based on the total flow demand, the electric control valve adjusts its opening based on the allocation ratio, and the electromagnetic directional valve switches the flow direction of the cooling circuit as needed. The actuators' movements are monitored in real time and under closed-loop control. The system uses feedback from flow, pressure, and temperature sensors to adjust control signals and compensate for execution errors. Furthermore, the system coordinates the cooling intensity of adjacent zones based on a heat transfer matrix to avoid localized overcooling or overheating. For example, if an abnormally high temperature is detected in a zone, the system not only increases the cooling flow in that zone but also strengthens cooling in upstream zones, the primary source of heat. Collaborative control employs a prediction-correction strategy, predicting temperature trends in each zone based on a heat transfer model and adjusting cooling parameters in advance to prevent temperature overshoots. Through this sophisticated flow and direction control, the system achieves precise temperature management of each compressor component, ensuring that key components operate within the optimal temperature range. This avoids wasted cooling resources and unnecessary temperature fluctuations, extending equipment life and reducing energy consumption.

[0093] Furthermore, the coolant flow regulating valve and flow direction control valve in the heat dissipation system adjust the coolant parameters of each control sub-area in real time according to the cooling resource allocation strategy, and at the same time, coordinately control the cooling intensity of adjacent areas based on the heat transfer matrix, including: prioritizing each control sub-area according to the heat load index, allocating the control cycle and response order according to the sorting result, and obtaining a control priority list; performing proportional-integral-differential control calculation on the flow parameters and flow direction parameters in the cooling resource allocation strategy, converting the control signal into a valve opening instruction, and obtaining a control sequence of the actuator; calculating the thermal coupling influence coefficient between each control sub-area based on the heat transfer matrix, adjusting the cooling resource allocation of adjacent areas according to the thermal coupling influence coefficient, and obtaining a coordinated control compensation amount; fusing the control sequence and the coordinated control compensation amount to generate a valve control instruction, and performing real-time closed-loop correction according to the dynamic response of the system to obtain a control signal for the coolant flow regulating valve and the flow direction control valve; applying the control signal to the coolant flow regulating valve and the flow direction control valve to adjust the coolant parameters of each control sub-area in real time.

[0094] Specifically, the heat load index, as a comprehensive indicator to measure the urgency of regional heat dissipation, directly determines the allocation priority of cooling resources. The sorting process first normalizes the heat load index to eliminate the differences in the magnitude of the index under different working conditions. The normalization uses the Min-Max method to map the heat load index of each region to the [0,1] interval. The system then sets three priority thresholds: 0.8, 0.5, and 0.3, and divides the control sub-areas into four priority levels: emergency cooling area (index>0.8), priority cooling area (0.5<index≤0.8), conventional cooling area (0.3<index≤0.5), and monitoring area (index≤0.3). Within each priority level, detailed sorting is performed based on the precise value of the heat load index. Based on the priority sorting results, the system assigns different control cycles and response orders. The emergency cooling zone uses a fast 100ms control cycle, with control commands executed first. The priority cooling zone uses a 200ms control cycle, with next-highest priority. The conventional cooling zone uses a 500ms control cycle. The monitoring zone only collects data without active regulation, with a control cycle of 1000ms. This hierarchical control approach ensures that computing and control resources are focused on the areas that require the most attention, while also ensuring the real-time performance of the overall system. The sorting results ultimately form a control priority list, which includes each zone's ID, priority level, control cycle, and execution order. This list guides the generation and execution of subsequent control signals.

[0095] Specifically, performing proportional-integral-derivative (PID) control calculations on the flow and direction parameters in the cooling resource allocation strategy is a key step in converting high-level strategies into specific execution instructions. This process employs a hierarchical PID control structure, consisting of a main-loop PID controller and multiple sub-loop PID controllers. The main-loop controller takes the deviation between the zone average temperature and the target temperature as input and outputs the total cooling capacity required for the zone. The sub-loop controllers decompose the total cooling capacity into specific control variables for the flow and direction parameters. For the flow parameter, the PID controller's proportional coefficient Kp is set between 0.8 and 1.2 (dynamically adjusted based on zone characteristics), the integral coefficient Ki is set between 0.1 and 0.3, and the differential coefficient Kd is set between 0.05 and 0.15. A large proportional coefficient ensures rapid system response to temperature changes, an appropriate integral coefficient eliminates steady-state errors, and a small differential coefficient provides necessary lead correction. The flow direction parameter is controlled using a discrete state switching approach, with the system determining the switching timing based on temperature distribution. Deadband and hysteresis mechanisms are introduced during the switching process to prevent system instability caused by frequent switching. PID calculations incorporate various anti-saturation measures, including integral limiting, variable-speed integral, and integral separation, effectively preventing integral saturation. The output of the control calculations must be converted into opening commands that the actuator can understand. This conversion process accounts for the nonlinear characteristics of the valves, ensuring a linear relationship between the controlled variable and the actual flow rate through linearization compensation. The resulting actuator control sequence contains opening commands for all regulating valves and directional valves within each control sub-area, along with corresponding execution timestamps. These commands are sent directly to the actuator to complete the specific operation.

[0096] Specifically, calculating the thermal coupling influence coefficient between each control sub-region based on the heat transfer matrix is the theoretical basis for achieving inter-regional coordinated control. The thermal coupling influence coefficient reflects the degree to which temperature changes in one region affect another. The calculation process is based on the heat transfer matrix and temperature response characteristics. The system first normalizes the heat transfer matrix to obtain the temperature response function for unit heat transfer. The thermal coupling influence coefficient is then calculated using the temperature response function and the time constant. This coefficient is an N×N matrix (N is the number of control sub-regions), and each element Hij represents the intensity of the impact of the temperature change in region i on region j. The calculation of the influence coefficient takes into account the time lag effect, that is, it takes a certain amount of time for heat transfer to be reflected in the temperature change. The time lag is described by a first-order lag model, and the time constant is determined by the physical distance and material properties between the regions, typically in the range of 10-300 seconds. Based on the thermal coupling influence coefficient, the system adjusts the cooling resource allocation between adjacent regions. The adjustment principle is: if area A has a strong thermal coupling effect on area B, and the temperature of area B is close to the upper limit, then increase the cooling intensity of area A; if areas A and B have a thermal impact on each other, coordinate the cooling timing of the two areas to avoid strengthening or weakening the cooling at the same time. This collaborative control based on thermal coupling takes into account the overall thermal balance of the system, rather than just focusing on the local control of a single area. The calculation of the collaborative control compensation amount adopts a weighted summation method, which comprehensively considers the temperature state, thermal coupling strength and time lag of each adjacent area to obtain a correction value for the original control amount. These compensation amounts ensure that the system will not cause new hot spots or waste energy while dealing with hot spot problems.

[0097] Specifically, the fusion of the control sequence and the coordinated control compensation is the process of generating the final control command. This process employs a combination of weighted fusion and priority arbitration. For continuous control variables, such as flow parameters, weighted fusion combines the original control sequence value and the coordinated control compensation in a 6:4 ratio to produce the fused control value. This ratio setting ensures the dominance of the original control strategy while providing sufficient adjustment space for coordinated control. For discrete control variables, such as flow direction switching, a priority arbitration method is employed. Specifically, when the coordinated control requirement has a higher priority than the original control sequence, the coordinated control command is adopted; otherwise, the original control sequence remains unchanged. The fused control command is then smoothed to avoid sudden changes in the control variable. This smoothing process utilizes a combination of sliding window averaging and rate-of-change limiting. The window size is dynamically adjusted based on the control cycle, while the rate-of-change limit is set based on the actuator's responsiveness. The processed command is converted into a standard valve control command, consisting of a position command, a velocity command, and an acceleration command, ensuring smooth valve movement. The control system performs closed-loop corrections through a real-time feedback loop, collecting feedback signals from flow, pressure, and temperature sensors, calculating control deviations, and updating control parameters. Closed-loop correction employs model predictive control (MPC) methods, leveraging a system model to predict system responses over a period of time (typically 5-10 control cycles) in the future, optimizing current control decisions. This combined feedforward and feedback control architecture improves the system's ability to suppress disturbances and enhances control accuracy. The resulting control signal, containing complete timing and control quantity information, directly drives the movement of the coolant flow control valve and the flow direction control valve.

[0098] Specifically, applying control signals to the coolant flow control valve and flow direction control valve is the final step in implementing heat dissipation control. Control signals are sent to each actuator via an industrial fieldbus such as Profibus-DP or Modbus. This process requires consideration of communication latency and signal synchronization. The system uses time synchronization protocols such as PTP (Precision Time Protocol) to ensure clock consistency across all actuators. Control instructions are timestamped, and the actuators use these timestamps to determine execution timing, ensuring coordinated operation of multiple actuators. The flow control valves are electric proportional valves driven by stepper motors or servo motors, achieving control accuracy of 1% of the valve's full stroke and a response time of less than 200ms. The flow direction control valves, using solenoid valves or electric ball valves, switch the flow direction of the cooling circuit. All valves are equipped with position feedback devices that report their current opening in real time, forming a complete closed-loop control system. To protect the actuators, the system implements multiple safety measures, including valve jam detection, overcurrent protection, and emergency shutdown logic. Stuck valve detection is achieved by comparing the difference between the control signal and the position feedback signal. When the difference exceeds a threshold and lasts longer than a set value, the system determines that the valve is stuck, issues an alarm, and switches to the backup circuit. Overcurrent protection monitors the motor drive current to prevent valve overload. Emergency shutdown logic switches the system to a safe state when a serious anomaly is detected, returning all flow valves to a preset safe position (designed according to fail-safe principles). Through these sophisticated control and protection measures, the system can accurately and reliably adjust coolant parameters in each control sub-zone, achieving precise heat dissipation control and ensuring safe and efficient operation of the air compressor under various operating conditions.

[0099] In this embodiment, a multidimensional thermal data set of a fuel cell air compressor is collected in layers; features are extracted from the multidimensional thermal data set, and the extracted feature data set is input into a neural network model based on deep transfer learning to obtain a corresponding heat dissipation efficiency attenuation prediction model; a multi-objective optimization equation for the air compressor is established based on the multidimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and the multi-objective optimization equation is solved to obtain a heat dissipation control strategy set; the heat dissipation system of the air compressor is dynamically partitioned, and the flow rate and flow direction of the coolant in each area of the dynamic partition are coordinated and adjusted according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor. The present invention solves the problems of heat dissipation efficiency attenuation and uneven resource allocation caused by changes in coolant quality and integrated control in the prior art.

[0100] The above describes the heat dissipation method of the fuel cell air compressor in the embodiment of the present invention. The following describes the heat dissipation device of the fuel cell air compressor in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a heat dissipation device for a fuel cell air compressor includes:

[0101] The data acquisition module 201 is used to perform layered acquisition of direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor to obtain a multi-dimensional thermal data set of the air compressor;

[0102] The pattern recognition module 202 is used to extract features from the multidimensional thermal data set, input the extracted feature data set into a neural network model based on deep transfer learning, and obtain a corresponding heat dissipation efficiency attenuation prediction model;

[0103] a strategy generation module 203 for establishing a multi-objective optimization equation for the air compressor based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solving the multi-objective optimization equation to obtain a heat dissipation control strategy set;

[0104] The partition control module 204 is used to dynamically partition the heat dissipation system of the air compressor and coordinately adjust the flow rate and flow direction of the coolant in each area of the dynamic partition according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor.

[0105] In an embodiment of the present invention, the fuel cell air compressor heat dissipation device operates the above-mentioned fuel cell air compressor heat dissipation method, wherein the fuel cell air compressor heat dissipation device collects a multidimensional thermal data set of the fuel cell air compressor in layers; extracts features from the multidimensional thermal data set, and inputs the extracted feature data set into a neural network model based on deep transfer learning to obtain a corresponding heat dissipation efficiency attenuation prediction model; establishes a multi-objective optimization equation for the air compressor based on the multidimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solves the multi-objective optimization equation to obtain a heat dissipation control strategy set; dynamically partitions the heat dissipation system of the air compressor, and coordinately adjusts the flow rate and flow direction of the coolant in each dynamically partitioned area according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor. The present invention solves the problems of heat dissipation efficiency attenuation and uneven resource allocation caused by changes in coolant quality and integrated control in the prior art.

[0106] above Figure 2 The fuel cell air compressor heat dissipation device in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The fuel cell air compressor heat dissipation device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0107] Figure 3This is a schematic diagram of the structure of a fuel cell air compressor heat dissipation device provided in an embodiment of the present invention. The fuel cell air compressor heat dissipation device 300 may vary significantly depending on configuration or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors), memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) storing application programs 333 or data 332. The memory 320 and storage media 330 may be either transient or persistent storage. The program stored in the storage medium 330 may include one or more modules (not shown), each of which may include a series of instructions for operating on the fuel cell air compressor heat dissipation device 300. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, executing the series of instructions stored in the storage medium 330 on the fuel cell air compressor heat dissipation device 300 to implement the steps of the above-described fuel cell air compressor heat dissipation method.

[0108] The fuel cell air compressor heat dissipation device 300 may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3 The illustrated fuel cell air compressor heat dissipation device structure does not constitute a limitation on the fuel cell air compressor heat dissipation device provided by the present invention, and may include more or fewer components than illustrated, or a combination of certain components, or a different arrangement of components.

[0109] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0110] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0111] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. 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 method for heat dissipation of a fuel cell air compressor, characterized in that: The fuel cell air compressor heat dissipation method includes: The direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor are collected in layers to obtain a multidimensional thermal data set of the air compressor; Time domain feature extraction, frequency domain feature extraction, and time-frequency feature extraction are performed on each parameter in the multidimensional thermal data set to obtain target domain features; the heat dissipation efficiency attenuation pattern features of the existing air compressor in the preset source domain knowledge base are used as source domain features, and the feature mapping algorithm of the attention mechanism is used to adapt and transform the source domain features and the target domain features; the target domain features are fused with the adaptively transformed source domain features, and the fused features are dynamically weighted through a hierarchical attention network to obtain a feature data set; the feature data set is input into a neural network model based on deep transfer learning for training, and a heat dissipation efficiency attenuation prediction model is obtained by minimizing the distribution difference between the source domain features and the target domain features and the heat dissipation efficiency prediction error; Establishing a multi-objective optimization equation for the air compressor based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solving the multi-objective optimization equation to obtain a heat dissipation control strategy set; The heat dissipation system of the air compressor is dynamically partitioned, and the flow rate and flow direction of the coolant in each area of the dynamic partition are coordinated and adjusted according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor.

2. The fuel cell air compressor heat dissipation method according to claim 1, characterized in that: The direct thermal characteristic parameters, indirect thermal characteristic parameters and coolant quality characteristic parameters of the fuel cell air compressor are collected in layers to obtain a multi-dimensional thermal data set of the air compressor, including: Setting a first layer of data collection points at a heat exchange position where the coolant of the fuel cell air compressor flows through, to collect the direct thermal characteristic parameters, wherein the direct thermal characteristic parameters include temperature data, flow rate data, and pressure data; Setting a second layer of data collection points at the preset core components of the air compressor to collect the indirect thermal characteristic parameters, wherein the indirect thermal characteristic parameters include speed data, current data, and vibration data; A third layer of data collection points is set in the coolant circulation system of the air compressor's heat dissipation system to collect coolant quality characteristic parameters, including conductivity, pH value and turbidity data; Time synchronization and data preprocessing are performed on the direct thermal characteristic parameters, the indirect thermal characteristic parameters, and the coolant quality characteristic parameters to obtain a multidimensional thermal data set of the air compressor.

3. The fuel cell air compressor heat dissipation method according to claim 1, characterized in that: The establishing of a multi-objective optimization equation for the air compressor based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solving the multi-objective optimization equation to obtain a heat dissipation control strategy set includes: Establishing a thermodynamic model of the air compressor heat dissipation system according to the multidimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and constructing the multi-objective optimization equation based on the thermodynamic model; Establishing a set of constraints for the multi-objective optimization equation, wherein the set of constraints includes upper temperature limits for each component, coolant flow range constraints, pump power limit constraints, and dynamic constraints based on coolant quality characteristic parameters; Inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer to solve and obtain a Pareto optimal solution set; The current operating condition is determined according to the multidimensional thermal data set, and the Pareto optimal solution set is dynamically screened according to the current operating condition, and a corresponding heat dissipation control strategy set is generated according to the screened Pareto optimal solution set.

4. The fuel cell air compressor heat dissipation method according to claim 3, characterized in that: Inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer to solve and obtain a Pareto optimal solution set includes: Inputting the multi-objective optimization equation and the constraint condition set thereof into a preset multi-objective dung beetle optimizer, and mapping the objective function space of the multi-objective optimization equation into the decision variable space by the multi-objective dung beetle optimizer to obtain an initial solution population of the optimization problem; Calculating the objective function value for each solution in the initial solution population to obtain an evaluation value matrix of the solution, calculating the dominance relationship between any two solutions based on the evaluation value matrix, hierarchically sorting the solutions based on the dominance relationship, and obtaining a position update vector for each solution; Applying the position update vector to each solution in the initial solution population to generate an updated solution set, and performing constraint checking on the solutions in the updated solution set. When a solution in the updated solution set violates the set of constraints, correcting the position of the corresponding solution using a preset penalty function to obtain a candidate solution set that satisfies the constraints; Calculating the crowding distance of the candidate solution set, sorting the solutions with the same non-dominated level based on the crowding distance, and selecting high-quality solutions through the elite retention strategy to obtain the non-dominated solution set in the iterative process; According to the preset iterative termination condition, the optimization process of the multi-objective dung beetle optimizer is cyclically iterated. When the termination condition is met, the Pareto front is extracted from the final non-dominated solution set to obtain the Pareto optimal solution set.

5. The fuel cell air compressor heat dissipation method according to claim 1, characterized in that: Dynamically partitioning the heat dissipation system of the air compressor and collaboratively adjusting the flow rate and flow direction of the coolant in each area of the dynamic partition according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor includes: extracting temperature field distribution information according to direct thermal characteristic parameters in the multidimensional thermal data set, and calculating heat flow characteristics according to the direct thermal characteristic parameters and the indirect thermal characteristic parameters; Clustering the temperature field distribution information and heat flow characteristics using an adaptive K-means clustering algorithm, and dynamically partitioning the heat dissipation system of the air compressor based on the clustering results to obtain N control sub-areas, where N is an integer from 2 to 8; Calculating a heat load index for each of the control sub-regions based on the direct thermal characteristic parameters, and constructing an inter-region heat transfer matrix to characterize the heat transfer relationship between the control sub-regions; Determining the cooling resource allocation strategy for each of the control sub-areas using a hierarchical progressive control method according to the heat dissipation control strategy set, the heat load index, and the heat transfer matrix; The coolant flow regulating valve and flow direction control valve in the heat dissipation system are used to adjust the coolant parameters of each control sub-area in real time according to the cooling resource allocation strategy, and the cooling intensity of adjacent areas is collaboratively controlled based on the heat transfer matrix to achieve heat dissipation of the air compressor.

6. The fuel cell air compressor heat dissipation method according to claim 5, characterized in that: The coolant flow regulating valve and the flow direction control valve in the heat dissipation system are used to adjust the coolant parameters of each control sub-area in real time according to the cooling resource allocation strategy, and the cooling intensity of adjacent areas is collaboratively controlled based on the heat transfer matrix, including: Prioritizing each of the control sub-areas according to the heat load index, allocating a control cycle and a response order according to the ranking result, and obtaining a control priority list; Performing proportional-integral-differential control calculation on the flow parameters and flow direction parameters in the cooling resource allocation strategy, converting the control signal into a valve opening instruction, and obtaining a control sequence of the actuator; Calculating the thermal coupling influence coefficient between each of the control sub-regions based on the heat transfer matrix, and adjusting the cooling resource allocation of adjacent regions according to the thermal coupling influence coefficient to obtain a collaborative control compensation amount; The control sequence and the coordinated control compensation are integrated to generate valve control instructions, and real-time closed-loop correction is performed according to the dynamic response of the system to obtain control signals for the coolant flow control valve and the flow direction control valve; The control signal is applied to the coolant flow regulating valve and the flow direction control valve to adjust the coolant parameters of each of the control sub-areas in real time.

7. A heat dissipation device for a fuel cell air compressor, characterized in that: The fuel cell air compressor heat dissipation device includes: A data acquisition module is used to perform layered acquisition of direct thermal characteristic parameters, indirect thermal characteristic parameters, and coolant quality characteristic parameters of the fuel cell air compressor to obtain a multidimensional thermal data set of the air compressor; A pattern recognition module is used to perform time domain feature extraction, frequency domain feature extraction, and time-frequency feature extraction on each parameter in the multidimensional thermal data set to obtain target domain features; the heat dissipation efficiency attenuation pattern features of the existing air compressor in the preset source domain knowledge base are used as source domain features, and the feature mapping algorithm of the attention mechanism is used to adapt and transform the source domain features and the target domain features; the target domain features are fused with the adaptively transformed source domain features, and the fused features are dynamically weighted through a hierarchical attention network to obtain a feature data set; the feature data set is input into a neural network model based on deep transfer learning for training, and a heat dissipation efficiency attenuation prediction model is obtained by minimizing the distribution difference and heat dissipation efficiency prediction error between the source domain features and the target domain features; a strategy generation module, configured to establish a multi-objective optimization equation for the air compressor based on the multi-dimensional thermal data set and the heat dissipation efficiency attenuation prediction model, and solve the multi-objective optimization equation to obtain a heat dissipation control strategy set; The partition control module is used to dynamically partition the heat dissipation system of the air compressor and coordinately adjust the flow rate and flow direction of the coolant in each area of the dynamic partition according to the heat dissipation control strategy set to achieve heat dissipation of the air compressor.

8. A fuel cell air compressor heat dissipation device, characterized in that: The fuel cell air compressor heat dissipation device includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory to enable the fuel cell air compressor heat dissipation device to perform the steps of the fuel cell air compressor heat dissipation method according to any one of claims 1 to 6.

Images