A real-time state monitoring analysis method and system for an industrial computer host

Abstract

This invention relates to the field of computer technology, and more particularly to a method and system for real-time status monitoring and analysis of industrial computer mainframes. First, a dust impact coefficient is obtained, which characterizes the degree of influence of dust accumulation on the heat dissipation of the industrial computer mainframe. Then, based on the dust impact coefficient, a dynamic model of heat dissipation efficiency is established, and the real-time heat dissipation efficiency is obtained based on the dynamic model. Next, based on the real-time heat dissipation efficiency, a temperature monitoring threshold is dynamically adjusted to obtain a real-time temperature monitoring threshold. Finally, real-time status monitoring and analysis are performed based on the real-time temperature monitoring threshold. This invention quantifies the effect of environmental pollutants on heat dissipation efficiency through the dust impact coefficient, and the dynamic model of heat dissipation efficiency established based on this, along with the dynamic adjustment of the temperature monitoring threshold, solves the problem that existing real-time status monitoring and analysis methods for industrial computers cannot adaptively adjust dynamically.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method and system for real-time status monitoring and analysis of industrial computer mainframes. Background Technology

[0002] Industrial computers serve as core control units in manufacturing, energy systems, and process control, requiring their mainframes to operate under complex conditions for extended periods. Compared to ordinary commercial computers, industrial computers may need to withstand multiple physical shocks, including extreme temperature fluctuations, high humidity, dust pollution, strong electromagnetic interference, and mechanical vibration. Against this backdrop, real-time status analysis technology becomes crucial for ensuring the continuous operation of industrial control systems.

[0003] However, existing industrial computer condition monitoring systems are generally based on fixed threshold matching algorithms, meaning that an alarm is only triggered when sensor data reaches a preset fixed threshold. This method has significant drawbacks; it cannot dynamically adapt to changes in equipment performance and the environment, often resulting in delayed warnings. For example, dust accumulation can reduce heat dissipation efficiency, causing equipment temperatures to rise more rapidly. If the fixed temperature threshold is not updated in time, even if the threshold is reached, CPU overheating and system crashes may occur due to reduced response time from maintenance personnel or other emergency response mechanisms.

[0004] Therefore, there is a need for a method that can utilize dynamic adaptive thresholds for real-time status monitoring and analysis of industrial computer mainframes. Summary of the Invention

[0005] Therefore, the present invention provides a real-time status monitoring and analysis method and system for industrial computer mainframes, in order to solve the problem that existing real-time status monitoring and analysis methods for industrial computers cannot adaptively and dynamically adjust.

[0006] This invention provides a real-time status monitoring and analysis method for industrial computer mainframes, comprising:

[0007] Obtain the dust impact coefficient, which is used to characterize the degree of impact of dust accumulation on the heat dissipation of industrial computer mainframes;

[0008] Based on the dust impact coefficient, a dynamic model of heat dissipation efficiency is established, and the real-time heat dissipation efficiency is obtained based on the dynamic model of heat dissipation efficiency.

[0009] Based on the real-time heat dissipation efficiency, the temperature monitoring threshold is dynamically adjusted to obtain the real-time temperature monitoring threshold.

[0010] Real-time status monitoring and analysis are performed based on real-time temperature monitoring thresholds.

[0011] In a preferred embodiment: the dust impact coefficient is obtained, including:

[0012] Obtain environmental parameters and hardware status parameters;

[0013] The dust impact coefficient is obtained based on environmental parameters and hardware status parameters.

[0014] In a preferred embodiment: the dust impact coefficient is obtained based on environmental parameters and hardware status parameters, including:

[0015] Based on environmental and hardware parameters, establish an input vector sequence;

[0016] The input vector sequence is input into the preset prediction model to obtain the dust influence coefficient output by the preset prediction model.

[0017] In a preferred embodiment: the input vector sequence includes an environmental parameter vector sequence and a hardware parameter vector sequence; the preset prediction model includes an input layer, a first prediction layer, a second prediction layer, a feature concatenation layer, a fully connected layer, and an output layer. The input layer connects the first and second prediction layers. The first prediction layer outputs a first feature vector based on the environmental parameter vector sequence, and the second prediction layer outputs a second feature vector based on the hardware parameter vector sequence. The first and second prediction layers connect to the feature concatenation layer, which concatenates the first and second feature vectors to obtain an intermediate vector. The feature concatenation layer connects to the fully connected layer, which connects to the output layer. The fully connected layer inputs the intermediate vector and outputs the dust influence coefficient through the output layer.

[0018] In a preferred embodiment: the preset prediction model is also used to output the predicted heat dissipation efficiency; the loss function of the preset prediction model is constructed by weighted summation of the first loss and the second loss, wherein the first loss represents the deviation of the output dust influence coefficient and the second loss represents the deviation of the output predicted heat dissipation efficiency.

[0019] In a preferred embodiment: the environmental parameters include at least one of dust concentration, temperature, humidity and air pressure, and the hardware status parameters include at least one of fan speed, fan voltage drop and CPU load.

[0020] In a preferred embodiment: a dynamic model of heat dissipation efficiency is established based on the dust influence coefficient, and the real-time heat dissipation efficiency is obtained based on the dynamic model, including:

[0021] Obtain the heat dissipation efficiency value and preset airflow blockage coefficient under ideal conditions;

[0022] A dynamic model of heat dissipation efficiency is established based on the heat dissipation efficiency value under ideal conditions, the preset air duct blockage coefficient, and the dust influence coefficient.

[0023] The real-time dust concentration and duct pressure drop are obtained, and the real-time heat dissipation efficiency is obtained based on the dynamic heat dissipation efficiency model.

[0024] In a preferred embodiment: the temperature monitoring threshold is dynamically adjusted based on the real-time heat dissipation efficiency to obtain the real-time temperature monitoring threshold, including:

[0025] Acquire the basic temperature monitoring threshold, reference temperature, real-time CPU load power, and preset ambient temperature compensation coefficient;

[0026] Obtain the ambient temperature and calculate the temperature deviation based on the difference between the ambient temperature and the reference temperature;

[0027] The real-time temperature monitoring threshold is obtained based on the basic temperature monitoring threshold, CPU real-time load power, preset ambient temperature compensation coefficient, temperature deviation, and real-time heat dissipation efficiency.

[0028] The present invention also provides a real-time status monitoring and analysis system for an industrial computer host, comprising:

[0029] The dust analysis module is used to obtain the dust impact coefficient, which characterizes the degree of impact of dust accumulation on the heat dissipation of the industrial computer host.

[0030] The model building module is used to build a dynamic model of heat dissipation efficiency based on the dust influence coefficient, and to obtain the real-time heat dissipation efficiency based on the dynamic model of heat dissipation efficiency.

[0031] The dynamic adjustment module is used to dynamically adjust the temperature monitoring threshold based on the real-time heat dissipation efficiency to obtain the real-time temperature monitoring threshold.

[0032] The monitoring and analysis module is used to perform real-time status monitoring and analysis based on real-time temperature monitoring thresholds.

[0033] The present invention also provides a computer-readable storage medium for storing a computer-readable program or instruction, which, when executed by a processor, can implement the steps in the real-time status monitoring and analysis method for an industrial computer host as described above.

[0034] The beneficial effects of using the above embodiments are:

[0035] This invention provides a real-time status monitoring and analysis method for industrial computer mainframes. First, a dust impact coefficient is obtained, which characterizes the degree of influence of dust accumulation on the heat dissipation of the industrial computer mainframe. Then, based on the dust impact coefficient, a dynamic model of heat dissipation efficiency is established, and the real-time heat dissipation efficiency is obtained according to the dynamic model. Next, the temperature monitoring threshold is dynamically adjusted based on the real-time heat dissipation efficiency to obtain the real-time temperature monitoring threshold. Finally, real-time status monitoring and analysis are performed based on the real-time temperature monitoring threshold. The real-time status monitoring and analysis method proposed in this invention achieves adaptive optimization of thermal management for industrial computers by constructing a closed-loop feedback mechanism between the dust impact coefficient and the dynamic model of heat dissipation efficiency. Firstly, the dust impact coefficient quantifies the quantitative effect of environmental pollutants on heat dissipation efficiency, breaking through the limitations of nonlinear dust deposition effects into calculable dynamic parameters. Based on this, a dynamic model of heat dissipation efficiency is established, and the temperature monitoring threshold is dynamically adjusted, thereby advancing the temperature warning time window and effectively avoiding the technical defect of "threshold alarm equals system crash." This solves the problem that existing real-time status monitoring and analysis methods for industrial computers cannot adaptively adjust dynamically. Attached Figure Description

[0036] Figure 1 A flowchart illustrating the real-time status monitoring and analysis method for an industrial computer host provided by the present invention.

[0037] Figure 2 This is a schematic diagram of the structure of the preset prediction model in this invention;

[0038] Figure 3 This is a system architecture diagram of the real-time status monitoring and analysis system for industrial computer mainframes provided by the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] Combination Figure 1 As shown in the figure, a specific embodiment of the present invention discloses a real-time status monitoring and analysis method for an industrial computer host, comprising:

[0041] S101. Obtain the dust impact coefficient, which is used to characterize the degree of impact of dust accumulation on the heat dissipation of the industrial computer host.

[0042] S102. Based on the dust influence coefficient, establish a dynamic model of heat dissipation efficiency, and obtain the real-time heat dissipation efficiency based on the dynamic model of heat dissipation efficiency.

[0043] S103. Based on the real-time heat dissipation efficiency, the temperature monitoring threshold is dynamically adjusted to obtain the real-time temperature monitoring threshold.

[0044] S104. Perform real-time status monitoring and analysis based on the real-time temperature monitoring threshold.

[0045] The real-time status monitoring and analysis method proposed in this invention achieves adaptive optimization of thermal management for industrial computers by constructing a closed-loop feedback mechanism based on a dynamic model of dust impact coefficient and heat dissipation efficiency. Firstly, the impact coefficient of dust is used to quantify the effect of environmental pollutants on heat dissipation efficiency, transforming the nonlinear dust deposition effect into a calculable dynamic parameter. Based on this, a dynamic model of heat dissipation efficiency is established, and the temperature monitoring threshold is dynamically adjusted. This advances the temperature warning time window, effectively avoiding the technical defect of "threshold alarm equals system crash," and solving the problem that existing real-time status monitoring and analysis methods for industrial computers cannot adaptively adjust dynamically.

[0046] It is understandable that in the above process, the dust impact coefficient can be a pre-set hyperparameter or data obtained based on experience. Specifically, the dust impact coefficient characterizes the inhibitory effect of dust on heat dissipation. For example, the larger the dust impact coefficient value (which can actually be smaller and set flexibly according to the situation), the more significant the reduction in heat dissipation capacity due to dust. For another example, when the dust impact coefficient is 0.15, it means that for every 10 μg / m³ increase in dust concentration, the heat dissipation efficiency decreases by approximately 15%. The dust impact coefficient can dynamically reflect environmental changes. For example, in an ideal dust-free environment, the dust impact coefficient is zero, and the heat dissipation efficiency is unaffected. However, as dust accumulates, the dust impact coefficient value gradually increases, and the heat dissipation efficiency decreases exponentially.

[0047] The invention also provides a preferred embodiment, wherein step S101, obtaining the dust influence coefficient, specifically includes:

[0048] Obtain environmental parameters and hardware status parameters;

[0049] The dust impact coefficient is obtained based on environmental parameters and hardware status parameters.

[0050] The aforementioned environmental parameters refer to parameters related to the environment, used to directly reflect dust accumulation and environmental conditions. These parameters include at least one of dust concentration, temperature, humidity, and air pressure. Hardware status parameters refer to parameters of the industrial computer host's status, used to indirectly characterize the current operating state of the cooling system. These parameters include at least one of fan speed, fan voltage drop, and CPU load. This embodiment, based on existing solutions, further calculates the dust impact coefficient in real time using both environmental and hardware parameters, significantly improving the calculation accuracy and real-time performance of the dust impact coefficient. It is understood that the specific process of obtaining the dust impact coefficient based on the environmental and hardware status parameters can be implemented using any existing method, such as theoretical derivation or thermodynamic model simulation.

[0051] Furthermore, the present invention provides a preferred embodiment, wherein the above step of obtaining the dust impact coefficient based on environmental parameters and hardware status parameters specifically includes:

[0052] Based on environmental and hardware parameters, establish an input vector sequence;

[0053] The input vector sequence is input into the preset prediction model to obtain the dust influence coefficient output by the preset prediction model.

[0054] Compared to the previous approach, the biggest difference in this embodiment is the generation of the dust impact coefficient through a predictive model. Firstly, the pre-defined predictive model can construct a dynamic mapping relationship, utilizing deep learning algorithms to capture the nonlinear time-varying patterns between dust accumulation and multi-dimensional environmental parameters (such as temperature, humidity, and air pressure) and hardware conditions (such as fan aging and CPU load fluctuations), thereby enabling early prediction of future dust impacts. Secondly, compared to the lag inherent in real-time calculations that rely on immediate sampling data, the pre-defined predictive model can update the predicted values ​​periodically using a sliding window mechanism, further improving the foresight of the dust impact coefficient, increasing response speed, and expanding the response window. Furthermore, the predictive model can adaptively update parameter weights through online incremental learning. When hardware configurations or operating conditions change, there is no need to recalibrate the entire computing system, greatly improving the system's maintainability and scalability.

[0055] Specifically, in combination Figure 2As shown, the input vector sequence includes an environmental parameter vector sequence and a hardware parameter vector sequence; the preset prediction model includes an input layer, a first prediction layer, a second prediction layer, a feature concatenation layer, a fully connected layer, and an output layer. The input layer connects the first prediction layer and the second prediction layer. The first prediction layer outputs a first feature vector based on the environmental parameter vector sequence, and the second prediction layer outputs a second feature vector based on the hardware parameter vector sequence. The first and second prediction layers connect to the feature concatenation layer, which concatenates the first and second feature vectors to obtain an intermediate vector. The feature concatenation layer connects to the fully connected layer, which connects to the output layer. The fully connected layer inputs the intermediate vector and outputs the dust influence coefficient through the output layer.

[0056] In the above process, both the first and second prediction layers can be composed of any neural network units such as LSTM and GRU (all existing technologies that can be understood by those skilled in the art). The difference lies in that the first prediction layer is used to process time-series data such as PM2.5, temperature, and humidity to capture the long-term trend of dust accumulation, while the second prediction layer is used to process high-frequency hardware signals such as duct pressure drop and fan speed to capture transient disturbances. The feature splicing layer is used to splice the analysis results of the two layers and input them into the fully connected layer for analysis. This embodiment significantly improves the accuracy and adaptability of dust impact coefficient prediction through an innovative dual-branch neural network architecture.

[0057] Furthermore, in a preferred embodiment, the preset prediction model is also used to output the predicted heat dissipation efficiency; the loss function of the preset prediction model is constructed by weighted summation of a first loss and a second loss, wherein the first loss represents the deviation of the output dust influence coefficient, and the second loss represents the deviation of the output predicted heat dissipation efficiency.

[0058] For example, the loss function of a pre-defined prediction model can be expressed as:

[0059] ;

[0060] in, For loss function, The first loss, This is the second loss. and These are weighting coefficients, which can be adjusted according to the actual situation. For example, their values ​​can be 0.7 and 0.3 respectively.

[0061] It is understandable that, since the dust impact coefficient cannot be obtained intuitively in practice, while the heat dissipation efficiency can be obtained directly through experiments or theoretical calculations, this embodiment further introduces the predicted heat dissipation efficiency as an auxiliary label to assist in training the preset prediction model to infer the dust impact coefficient. This can provide the model with a clear convergence direction, avoid the cumulative error caused by the indirect derivation of the dust impact coefficient, and improve the accuracy of the preset prediction model.

[0062] Furthermore, in a preferred embodiment, step S102, establishing a dynamic model of heat dissipation efficiency based on the dust influence coefficient, and obtaining the real-time heat dissipation efficiency based on the dynamic model, specifically includes:

[0063] Obtain the heat dissipation efficiency value and preset airflow blockage coefficient under ideal conditions;

[0064] A dynamic model of heat dissipation efficiency is established based on the heat dissipation efficiency value under ideal conditions, the preset air duct blockage coefficient, and the dust influence coefficient.

[0065] The real-time dust concentration and duct pressure drop are obtained, and the real-time heat dissipation efficiency is obtained based on the dynamic heat dissipation efficiency model.

[0066] A feasible formula for the above process is as follows:

[0067] ;

[0068] in, This indicates the real-time heat dissipation efficiency. This represents the heat dissipation efficiency value under ideal conditions. Indicates the dust impact coefficient. This indicates the real-time dust concentration. This indicates the preset air duct blockage coefficient (which can be set based on the physical structure of the specific equipment and experience). This indicates the pressure drop in the air duct.

[0069] Furthermore, in a preferred embodiment, step S103, dynamically adjusting the temperature monitoring threshold based on the real-time heat dissipation efficiency to obtain the real-time temperature monitoring threshold, specifically includes:

[0070] Acquire the basic temperature monitoring threshold, reference temperature, real-time CPU load power, and preset ambient temperature compensation coefficient;

[0071] Obtain the ambient temperature and calculate the temperature deviation based on the difference between the ambient temperature and the reference temperature;

[0072] The real-time temperature monitoring threshold is obtained based on the basic temperature monitoring threshold, CPU real-time load power, preset ambient temperature compensation coefficient, temperature deviation, and real-time heat dissipation efficiency.

[0073] A feasible formula for the above process is as follows:

[0074] ;

[0075] in, Indicates the real-time temperature monitoring threshold. Indicates the baseline temperature monitoring threshold. Indicates the real-time CPU load power. This indicates the preset ambient temperature compensation coefficient (which can also be set based on experience). This indicates temperature deviation.

[0076] Combination Figure 3 As shown, the present invention also provides a real-time status monitoring and analysis system for an industrial computer host, comprising:

[0077] The dust analysis module 310 is used to obtain the dust influence coefficient, which is used to characterize the degree of influence of dust accumulation on the heat dissipation of the industrial computer host.

[0078] The model building module 320 is used to build a dynamic model of heat dissipation efficiency based on the dust influence coefficient, and to obtain the real-time heat dissipation efficiency based on the dynamic model of heat dissipation efficiency.

[0079] The dynamic adjustment module 330 is used to dynamically adjust the temperature monitoring threshold according to the real-time heat dissipation efficiency to obtain the real-time temperature monitoring threshold.

[0080] The monitoring and analysis module 340 is used to perform real-time status monitoring and analysis based on the real-time temperature monitoring threshold.

[0081] It should be noted that the corresponding systems provided in the above embodiments can implement the technical solutions described in the above method embodiments. The specific implementation principles of the above modules or units can be found in the corresponding content in the above method embodiments, and will not be repeated here.

[0082] This embodiment also provides a computer-readable storage medium storing a real-time status monitoring and analysis program for an industrial computer host. When the real-time status monitoring and analysis program for the industrial computer host is executed by a processor, it can implement the steps in the above embodiments.

[0083] This invention provides a real-time status monitoring and analysis method for industrial computer mainframes. First, a dust impact coefficient is obtained, which characterizes the degree of influence of dust accumulation on the heat dissipation of the industrial computer mainframe. Then, based on the dust impact coefficient, a dynamic model of heat dissipation efficiency is established, and the real-time heat dissipation efficiency is obtained according to the dynamic model. Next, the temperature monitoring threshold is dynamically adjusted based on the real-time heat dissipation efficiency to obtain the real-time temperature monitoring threshold. Finally, real-time status monitoring and analysis are performed based on the real-time temperature monitoring threshold. The real-time status monitoring and analysis method proposed in this invention achieves adaptive optimization of thermal management for industrial computers by constructing a closed-loop feedback mechanism between the dust impact coefficient and the dynamic model of heat dissipation efficiency. Firstly, the dust impact coefficient quantifies the quantitative effect of environmental pollutants on heat dissipation efficiency, breaking through the limitations of nonlinear dust deposition effects into calculable dynamic parameters. Based on this, a dynamic model of heat dissipation efficiency is established, and the temperature monitoring threshold is dynamically adjusted, thereby advancing the temperature warning time window and effectively avoiding the technical defect of "threshold alarm equals system crash." This solves the problem that existing real-time status monitoring and analysis methods for industrial computers cannot adaptively adjust dynamically.

[0084] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0085] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for real-time status monitoring and analysis of an industrial computer mainframe, characterized in that, include: Obtain environmental parameters and hardware status parameters; Based on environmental parameters and hardware status parameters, an input vector sequence is established; The input vector sequence is fed into a preset prediction model to obtain the dust impact coefficient output by the preset prediction model. The dust impact coefficient characterizes the degree of influence of dust accumulation on the heat dissipation of the industrial computer host. The input vector sequence includes an environmental parameter vector sequence and a hardware state parameter vector sequence. The preset prediction model includes an input layer, a first prediction layer, a second prediction layer, a feature concatenation layer, a fully connected layer, and an output layer. The input layer connects the first and second prediction layers. The first prediction layer outputs a first feature vector based on the environmental parameter vector sequence, and the second prediction layer outputs a second feature vector based on the hardware state parameter vector sequence. The first and second prediction layers connect to the feature concatenation layer, which concatenates the first and second feature vectors to obtain an intermediate vector. The feature concatenation layer connects to the fully connected layer, which connects to the output layer. The fully connected layer inputs the intermediate vector and outputs the dust impact coefficient through the output layer. The preset prediction model also outputs the predicted heat dissipation efficiency. The loss function of the preset prediction model is constructed by a weighted sum of the first loss and the second loss. The first loss represents the deviation of the output dust impact coefficient, and the second loss represents the deviation of the output predicted heat dissipation efficiency. Based on the dust impact coefficient, a dynamic model of heat dissipation efficiency is established, and the real-time heat dissipation efficiency is obtained based on the dynamic model of heat dissipation efficiency. Based on the real-time heat dissipation efficiency, the temperature monitoring threshold is dynamically adjusted to obtain the real-time temperature monitoring threshold. Real-time status monitoring and analysis are performed based on real-time temperature monitoring thresholds.

2. The real-time status monitoring and analysis method for industrial computer mainframes according to claim 1, characterized in that, Environmental parameters include at least one of dust concentration, temperature, humidity and air pressure, and hardware status parameters include at least one of fan speed, fan voltage drop and CPU load.

3. The real-time status monitoring and analysis method for industrial computer mainframes according to claim 1, characterized in that, Based on the dust impact coefficient, a dynamic model of heat dissipation efficiency is established, and the real-time heat dissipation efficiency is obtained based on the dynamic model, including: Obtain the heat dissipation efficiency value and preset airflow blockage coefficient under ideal conditions; A dynamic model of heat dissipation efficiency is established based on the heat dissipation efficiency value under ideal conditions, the preset air duct blockage coefficient, and the dust influence coefficient. The real-time dust concentration and duct pressure drop are obtained, and the real-time heat dissipation efficiency is obtained based on the dynamic heat dissipation efficiency model.

4. The real-time status monitoring and analysis method for industrial computer mainframes according to claim 1, characterized in that, Based on the real-time heat dissipation efficiency, the temperature monitoring threshold is dynamically adjusted to obtain the real-time temperature monitoring threshold, including: Acquire the basic temperature monitoring threshold, reference temperature, real-time CPU load power, and preset ambient temperature compensation coefficient; Obtain the ambient temperature and calculate the temperature deviation based on the difference between the ambient temperature and the reference temperature; The real-time temperature monitoring threshold is obtained based on the basic temperature monitoring threshold, CPU real-time load power, preset ambient temperature compensation coefficient, temperature deviation, and real-time heat dissipation efficiency.

5. A real-time status monitoring and analysis system for an industrial computer mainframe, characterized in that, include: The dust analysis module is used to acquire environmental parameters and hardware status parameters; Based on environmental parameters and hardware status parameters, an input vector sequence is established; The input vector sequence is fed into a preset prediction model to obtain the dust impact coefficient output by the preset prediction model. The dust impact coefficient characterizes the degree of influence of dust accumulation on the heat dissipation of the industrial computer host. The input vector sequence includes an environmental parameter vector sequence and a hardware state parameter vector sequence. The preset prediction model includes an input layer, a first prediction layer, a second prediction layer, a feature concatenation layer, a fully connected layer, and an output layer. The input layer connects the first and second prediction layers. The first prediction layer outputs a first feature vector based on the environmental parameter vector sequence, and the second prediction layer outputs a second feature vector based on the hardware state parameter vector sequence. The first and second prediction layers connect to the feature concatenation layer, which concatenates the first and second feature vectors to obtain an intermediate vector. The feature concatenation layer connects to the fully connected layer, which connects to the output layer. The fully connected layer inputs the intermediate vector and outputs the dust impact coefficient through the output layer. The preset prediction model also outputs the predicted heat dissipation efficiency. The loss function of the preset prediction model is constructed by a weighted sum of the first loss and the second loss. The first loss represents the deviation of the output dust impact coefficient, and the second loss represents the deviation of the output predicted heat dissipation efficiency. The model building module is used to build a dynamic model of heat dissipation efficiency based on the dust influence coefficient, and to obtain the real-time heat dissipation efficiency based on the dynamic model of heat dissipation efficiency. The dynamic adjustment module is used to dynamically adjust the temperature monitoring threshold based on the real-time heat dissipation efficiency to obtain the real-time temperature monitoring threshold. The monitoring and analysis module is used to perform real-time status monitoring and analysis based on real-time temperature monitoring thresholds.

6. A computer-readable storage medium, characterized in that, Used to store computer-readable programs or instructions, which, when executed by a processor, are capable of implementing the steps in any of the above claims 1-4 for the real-time status monitoring and analysis method for an industrial computer host.

Images