Constant-temperature machine case environment sensing intelligent regulation method and system

Abstract

This invention relates to the field of environmental sensing and intelligent control technology, specifically disclosing a method and system for intelligent control of a constant-temperature enclosure based on environmental sensing. The method includes: collecting temperature, humidity, and condensation area data within the enclosure; extracting condensation trend factors to generate a state initialization structure; constructing a thermo-humidity coupled state modeling structure; calculating the entropy increase sequence and humidity critical point to generate a set of control mapping parameters; constructing a power allocation strategy based on this parameter set to generate a set of execution control commands; controlling the state of the hydrophilic guide channel after determining condensate generation conditions and establishing a re-evaporation path configuration structure; collecting the microenvironment response sequence during execution and generating a feedback acquisition structure; and correcting the state modeling structure and updating the control parameters based on the feedback structure. This invention enables early prediction, proactive intervention, and dynamic adaptive updating of control precision for condensation within a sealed enclosure, improving the long-term reliability and anti-condensation stability of the enclosure system.

Description

Technical Field

[0001] This invention relates to the field of environmental sensing and intelligent control technology, and in particular to a method and system for intelligent control of constant temperature chassis environment sensing. Background Technology

[0002] With the widespread deployment of highly integrated and precision electronic equipment, highly sealed enclosures are commonly used to prevent dust, water, and corrosion, ensuring long-term stable operation under harsh sealing conditions. However, in such sealed structures, traditional temperature control systems often only perform closed-loop regulation based on a single temperature dimension, failing to address the safety hazards posed by humidity accumulation and condensation. Continuous humidity buildup inside the enclosure can easily induce circuit board corrosion, component short circuits, and other failures, especially in scenarios with significant temperature variations or high ambient humidity, where condensation is more likely to occur, severely impacting the equipment's operational reliability and lifespan.

[0003] Existing solutions attempt to incorporate dehumidifying fans or desiccant components, but they still lack effective condensation prediction mechanisms and multi-source information-driven control models, making it difficult to achieve early detection and proactive defense control of condensation risks. Furthermore, as a strongly nonlinear coupled system, the humid and thermal environment exhibits a high degree of temporal correlation between temperature and humidity changes; modeling temperature or humidity alone often fails to accurately predict condensation trends. Traditional methods also struggle to fully express the dynamic evolution characteristics of the chassis microenvironment in control strategies, resulting in lag and poor adaptability in control system responses.

[0004] In addition, existing solutions generally do not form a closed-loop intelligent control system from environmental perception, state modeling, strategy execution to error feedback correction, and lack efficient and reliable feedback self-learning capabilities. The control system cannot dynamically optimize parameter configuration based on actual operating data, resulting in problems such as decreased control accuracy and low energy efficiency during long-term operation.

[0005] In summary, there is an urgent need for an intelligent control method that can integrate temperature and humidity data with condensation trend factors, construct a thermo-humidity coupled state modeling structure based on entropy increase sequences, and dynamically generate humidity critical points and multi-dimensional control parameters. This method would build a closed-loop system integrating condensation prediction, active control, and feedback updates to achieve high-precision, high-response, and high-adaptive control of the microenvironment inside a sealed enclosure, meeting the safety and stability requirements of the next generation of highly reliable electronic equipment. Summary of the Invention

[0006] This invention provides a method and system for intelligent control of a constant temperature enclosure environment, which solves the problem of how to construct a thermo-humidity coupled state modeling structure with entropy increase sequence driving capability based on temperature and humidity data and condensation trend factors, dynamically generate humidity critical points and output a set of control mapping parameters, and realize an integrated intelligent control closed loop of condensation prediction, active control and adaptive feedback correction in a sealed enclosure.

[0007] To address the aforementioned technical problems, this invention provides a method for intelligent control of a constant-temperature chassis environment, comprising:

[0008] Acquire temperature and humidity data and condensation area data, extract condensation trend factors and fuse them to generate a state initialization structure; the temperature and humidity include instantaneous temperature values, humidity values ​​and time series records of multiple sampling nodes; the condensation area data includes local condensation state identifier values ​​collected by multiple condensation sensors;

[0009] Obtain the state initialization structure, construct the thermal-humidity coupled state modeling structure, calculate the entropy increase sequence and humidity critical point, and generate a set of control mapping parameters;

[0010] The expression for the thermal-humid coupling state modeling structure is:

[0011]

[0012] Among them, S k+1 This represents the thermal-humid coupling state structure at the (k+1)th time step; This represents the state prediction function based on a Long Short-Term Memory (LSTM) neural network. Indicates the state S k The nonlinear correction function after correcting for the physical driving term and the disturbance term; h k ,c k The hidden state and cellular state of LSTM;

[0013] The expression for the set of control mapping parameters is:

[0014]

[0015] Among them, C k To control the set of mapping parameters; This refers to the power of the heating element; This refers to the dehumidifier fan speed; This refers to the current of the thermoelectric cooler. Here is the control parameter mapping function; k is the time window number; This represents the rate of increase of entropy during the k-th time period; This represents the dynamic humidity critical point at time k.

[0016] Based on the set of control mapping parameters, a power allocation strategy and resource scheduling structure are constructed, data integration processing is performed, and a set of execution control instructions is generated.

[0017] Based on the humidity critical point and condensation trend factor, the conditions for condensate formation are determined, the state of the hydrophilic guide channel is controlled, and a re-evaporation path configuration structure is established.

[0018] The set of execution control instructions and the re-evaporation path configuration structure are obtained. Temperature and humidity changes are collected and microenvironment response sequences are generated. After calculating the error and analyzing the control response, a feedback acquisition structure is generated.

[0019] The thermal-humid coupling state modeling structure is corrected based on the feedback acquisition structure and the set of control mapping parameters, and then updated to the state initialization structure.

[0020] Furthermore, the steps of acquiring temperature and humidity data and condensation area data, extracting condensation trend factors, and fusing them to generate a state initialization structure include:

[0021] Acquire temperature, humidity, and condensation area data from inside the chassis, remove outliers, and standardize the format to obtain microenvironment data;

[0022] Extract temperature and humidity change curves, local humidity gradients, and condensation area characteristics from microenvironment data, and calculate condensation trend factors;

[0023] The condensation trend factor is fused with microenvironment data to generate a state initialization structure.

[0024] Furthermore, the construction of the thermal-humidity coupled state modeling structure includes:

[0025] Obtain the state initialization structure and construct the thermal-humidity coupled state modeling structure.

[0026] Furthermore, the calculation of the entropy increase sequence and the humidity critical point includes:

[0027] Based on the thermal-humidity coupled state modeling structure, an entropy increase sequence is constructed to represent the microenvironment change trend;

[0028] The humidity critical point is calculated based on the entropy increase sequence and is used to indicate the threshold for determining the occurrence of condensation.

[0029] Further, the generation of the control mapping parameter set includes:

[0030] The entropy increase sequence and humidity critical point are normalized to generate a set of control mapping parameters that include power, wind speed and current regulation information.

[0031] Furthermore, the construction of the power allocation strategy and resource scheduling structure includes:

[0032] Obtain the set of control mapping parameters and construct the power allocation strategy and resource scheduling structure;

[0033] The power allocation strategy and resource scheduling structure are calculated to obtain the heating element power, dehumidifying fan speed and cooling element current data.

[0034] The data on heating element power, dehumidifying fan speed, and cooling element current are integrated and processed to generate a set of execution control commands.

[0035] Furthermore, the structure for controlling the state of the hydrophilic guide channel and establishing the re-evaporation path configuration includes:

[0036] Obtain the humidity critical point and condensation tendency factor to determine the conditions for condensate formation;

[0037] The opening or closing status of the hydrophilic guide channel is controlled according to the condensate generation conditions;

[0038] Based on the state of the hydrophilic guide channel and the opening state of the evaporation membrane, a re-evaporation path configuration structure is established.

[0039] Furthermore, the step of collecting temperature and humidity changes and generating a microenvironment response sequence, calculating errors and analyzing control responses to generate a feedback acquisition structure includes:

[0040] Obtain the set of execution control commands and the re-evaporation path configuration structure, and jointly control the working status of the heating element, cooling element and dehumidifying fan;

[0041] During the control process, temperature and humidity change data are collected to generate a microenvironment response sequence.

[0042] Dynamic error calculation and control response analysis are performed on the microenvironment response sequence to generate a feedback acquisition structure.

[0043] Furthermore, the modification of the thermal-humid coupling state modeling structure and the updating to the state initialization structure include:

[0044] Obtain the feedback acquisition structure and control mapping parameter set, and construct a state mapping and control relationship table;

[0045] Based on the state mapping and control relationship table, the thermal-humid coupling state modeling structure and power allocation strategy parameters are corrected.

[0046] Update the modified thermal-humid coupling state modeling structure and power allocation strategy parameters to the state initialization structure.

[0047] A constant temperature chassis environment sensing and intelligent control system, applied to any of the above-described constant temperature chassis environment sensing and intelligent control methods, includes:

[0048] The environmental data acquisition module is used to acquire temperature and humidity data and condensation area data, and generate a state initialization structure;

[0049] The condensation trend analysis and state initialization module is used to extract temperature and humidity change curves, local humidity gradients, and condensation area characteristics from microenvironment data.

[0050] The modeling and entropy increase calculation module is used to construct a thermal-humidity coupled state modeling structure and calculate the state entropy increase sequence and humidity critical point based on the structure.

[0051] The regulation strategy generation and instruction construction module is used to construct power allocation strategies and resource scheduling structures based on the set of control mapping parameters, and to perform scheduling calculations accordingly.

[0052] The condensate recovery control module is used to determine the condensate generation conditions based on the humidity critical point and condensation trend factor, and control the state of the hydrophilic guide channel accordingly.

[0053] The execution feedback acquisition and parameter update module is used to obtain the set of execution control commands and the re-evaporation path configuration structure, jointly control the working status of the heating element, cooling element and dehumidifying fan, and collect the temperature and humidity changes inside the chassis during the control execution process to generate a microenvironment response sequence.

[0054] The key innovations of this invention include:

[0055] (1) Modeling structure of thermal and humid state driven by condensation trend factor and entropy increase sequence. The characteristics of condensation area are dynamically integrated with temperature and humidity gradient, and an entropy increase sequence is defined to reflect the change in environmental orderliness, forming a modeling foundation that is more predictive and generalizable than traditional temperature and humidity thresholds.

[0056] (2) Multi-execution unit joint control mechanism driven by control mapping parameter set. A control mapping parameter structure with normalized entropy growth rate and humidity critical point as the core is constructed to support the linkage control of heating, dehumidification and refrigeration equipment, and has the advantages of high control accuracy, flexible resource allocation and stable control convergence.

[0057] (3) Adaptive update mechanism linking re-evaporation path configuration structure and feedback acquisition structure. By performing error calculation and strategy inversion on the execution control results and microenvironment response sequence, the state modeling structure and control strategy parameters are updated in real time, ensuring the robustness and convergence of the system under various operating conditions.

[0058] The following are its main beneficial effects:

[0059] (1) Improve the response accuracy and predictive foresight of microenvironment perception. This invention overcomes the problems of delayed response and slow regulation of traditional systems to condensation generation state by introducing a joint modeling approach of "condensation trend factor" and "entropy increase sequence". The system uses time series features extracted from temperature and humidity data to dynamically calculate the rate of increase of state entropy, and combines the Navier-Stokes fluid model with thermal expansion.

[0060] (2) Achieving active and adjustable heat and humidity control and multi-source resource coordination. This invention innovatively constructs a "control mapping parameter set" based on normalized entropy increment and humidity critical point. This set is used to simultaneously drive the coordinated work of multiple execution units such as heating element power, dehumidifying fan speed and semiconductor cooling chip current. Compared with traditional temperature control systems that can only respond at a single point, this system can form a multi-dimensional control strategy combination according to different environmental disturbances, ensuring that the system can maintain a highly stable operating environment with temperature ±1℃ and humidity ≤40% under complex background of heat and humidity coupling.

[0061] (3) Construct a dynamic feedback-driven state modeling correction mechanism to enhance long-term adaptability and system robustness. During the execution of the control strategy, the system introduces a "micro-environment response sequence" and a "feedback acquisition structure" to dynamically calculate the error and backtrack the deviation between the actual temperature and humidity changes and the predicted state after execution. Then, through error weight remapping, the thermal-humidity coupling state modeling structure and control parameters are automatically corrected, which reduces the risk of failure caused by model drift and hardware aging during long-term system operation while maintaining control accuracy, thus forming a closed-loop control capability with "self-sensing - self-adjustment - self-adaptation". Attached Figure Description

[0062] Figure 1 A flowchart illustrating the intelligent control method for constant temperature chassis environment sensing provided in this application embodiment;

[0063] Figure 2 This is a structural block diagram of the constant temperature chassis environment sensing and intelligent control system provided in the embodiments of this application. Detailed Implementation

[0064] Example 1: Refer to Figure 1 This is a flowchart illustrating the intelligent control method for constant temperature chassis environment sensing provided in an embodiment of the present invention. The process may include at least steps S100-S600:

[0065] S100: Acquire temperature and humidity data and condensation area data, extract condensation trend factors and fuse them to generate a state initialization structure.

[0066] S200: Obtain the state initialization structure, construct the thermal-humidity coupled state modeling structure, calculate the entropy increase sequence and humidity critical point, and generate a set of control mapping parameters.

[0067] S300 constructs a power allocation strategy and resource scheduling structure based on the control mapping parameter set, performs data integration processing, and generates a set of execution control instructions.

[0068] S400: Based on the humidity critical point and condensation trend factor, determine the condensate generation conditions, control the state of the hydrophilic guide channel, and establish a re-evaporation path configuration structure.

[0069] S500: Acquire the set of execution control instructions and the re-evaporation path configuration structure, collect temperature and humidity changes and generate a microenvironment response sequence, calculate errors and analyze responses to generate a feedback acquisition structure.

[0070] S600: Correct the thermal-humid coupling state modeling structure based on the feedback acquisition structure and control mapping parameter set, and update it to the state initialization structure.

[0071] Step S100 includes at least steps S110-S130:

[0072] S110: Acquire temperature, humidity, and condensation area data within the chassis, remove outliers, and standardize the format to obtain microenvironment data.

[0073] In this step, the system collects temperature, humidity, and condensation zone data through a multi-point environmental sensor array deployed inside a sealed enclosure. The temperature and humidity data are collected in real time by high-precision digital temperature and humidity sensors, while the condensation zone data is obtained through local condensation status monitoring units located at condensation risk points.

[0074] Specifically, the temperature data includes instantaneous temperature values ​​and time-series records from multiple sampling nodes, covering micro-spaces at different heights, in different air ducts, and near device surfaces. The humidity data is acquired using the same method, forming a set of spatially synchronized humidity records corresponding to the temperature data. The condensation area data includes local condensation state identifier values ​​collected by multiple condensation sensors, and surface condensation phenomena are detected using methods such as conductivity abrupt changes, changes in surface light reflection, or micro-piezoelectric response.

[0075] After data acquisition, the system performs outlier removal processing on the temperature data, humidity data, and condensation area data. The outlier removal process includes static threshold removal, sliding window differential abrupt change removal, and time series trend smoothing. Static threshold removal is used to exclude recorded values ​​outside the physical range caused by sensor malfunctions; sliding window differential change removal is used to identify abrupt outliers; and time series trend smoothing is used to reduce fluctuations caused by random noise.

[0076] After outlier removal, the system performs format standardization on all data. Format standardization includes data structure reconstruction, sampling frequency alignment, and data field standardization. Specifically, this includes: remapping all temperature and humidity sampling results into a two-dimensional matrix structure under a unified sampling time axis; performing time interpolation matching operations on condensation area data synchronized with temperature and humidity data; and standardizing data units, field naming, and record structure to form a data structure with complete spatial and temporal consistency.

[0077] The temperature and humidity data, after being processed in a unified format, are combined with the condensation area data to construct a microenvironment data structure. This microenvironment data serves as the basis for calculating the condensation trend factor in subsequent step S120, and will also be fused with the condensation trend factor in step S130.

[0078] S120. Extract temperature and humidity change curves, local humidity gradients, and condensation area characteristics from microenvironment data, and calculate the condensation trend factor.

[0079] After acquiring microenvironment data, the system enters the condensation risk evolution trend modeling stage based on data structures. First, the system extracts temperature and humidity change curves from the microenvironment data. These curves are obtained by statistically analyzing the rate of temperature and humidity change at different sensing points within a time window. Further, the system constructs temperature and humidity change trajectory maps to represent the dynamic characteristics of heat flow distribution and moisture migration.

[0080] The system calculates the local humidity gradient based on the aforementioned temperature and humidity change curves. The local humidity gradient refers to the difference in the rate of humidity change between spatially adjacent sampling points, used to characterize the potential condensation migration trend in a micro-space. The system employs neighborhood difference calculation and spatial vectorization to establish a local gradient mapping map, which reflects the possible direction and regional distribution characteristics of condensed water vapor accumulation.

[0081] The system combines condensation region data to extract condensation region features. These features include statistical parameters such as the spatial distribution density of identified condensation locations, condensation frequency, and duration. The system performs clustering and morphological analysis on historical data of condensation regions to construct an evolution trend map of condensation regions and calculates the condensation activity index per unit time.

[0082] Based on temperature and humidity change curves, local humidity gradients, and condensation region characteristics, the system employs a fusion modeling approach to construct a condensation trend factor. This condensation trend factor is a comprehensive index describing the potential for condensation formation, and its construction process includes the following three steps: first, constructing a trend score based on a cross-scoring model of humidity increase rate and temperature decrease rate; second, generating a condensation accumulation potential distribution map based on the local humidity gradient; and third, correcting the aforementioned trend model based on the evolution trend of the condensation region to complete the generation of the final condensation trend factor.

[0083] The generated condensation trend factor will be used in the next step S130 to fuse with the microenvironment data to construct the state initialization structure, and will serve as the input basis for thermal-humidity coupling state modeling in the subsequent step S210.

[0084] S130. The condensation trend factor is fused with the microenvironment data to generate a state initialization structure.

[0085] After the condensation trend factor is constructed, the system enters the state initialization stage, which is specifically implemented by fusing the condensation trend factor with the microenvironment data to generate a state initialization structure.

[0086] Specifically, the fusion process includes two steps: structural layer fusion and semantic layer fusion. Structural layer fusion aligns the condensation trend factor vector with the microenvironment data structure by timestamp, constructing a multi-channel temporal feature matrix. Each channel includes a temperature sequence, humidity sequence, condensation region state sequence, and condensation trend factor sequence. This multi-channel structure is used for feature encoding and temporal inference in subsequent models.

[0087] Semantic layer fusion uses pattern recognition algorithms to correlate and express the interpretive relationship between condensation trend factors and microenvironment data. For example, the system uses a pattern labeling module to label typical temperature and humidity combinations corresponding to high condensation trend factor intervals, thereby assigning environmental labels, risk level labels, and control priority labels to the state structure.

[0088] The merged state initialization structure includes the following information fields: temperature distribution status, humidity distribution status, condensation area status, condensation trend level, multi-channel temperature and humidity time series matrix, local humidity gradient vector set, and condensation risk prediction label at the current time point.

[0089] The state initialization structure serves as the input basis for constructing the thermo-humidity coupled state modeling structure in subsequent step S210, ensuring that the calculation process of the entropy increase sequence and humidity critical point possesses environmental timeliness, locality, and predictive characteristics. In particular, in step S220, the condensation trend factor included in the state initialization structure will participate in the evaluation of the entropy increase dynamic characteristics and generate a set of control mapping parameters in subsequent step S230, providing a direct basis for the construction of the power allocation strategy and resource scheduling structure in S300.

[0090] Through the above implementation steps, the system realizes a full-link data processing flow from initial environmental perception to the initial state of microenvironment modeling. Compared with the traditional method of regulation based solely on static temperature and humidity data, this invention uses a condensation trend factor to predict condensation risk and integrates it with multi-dimensional microenvironment parameters to generate a state initialization structure, achieving dynamic modeling and prediction-driven control of the controlled object. Outlier removal and format standardization of the microenvironment data improve data reliability. The construction of the condensation trend factor introduces a comprehensive ability to express spatial gradient changes and regional evolution behavior. The output of the state initialization structure provides a real-time and predictive parameter basis for subsequent thermo-humidity coupled state modeling, enhancing the continuity, stability, and foresight of the system in performing regulation tasks in complex closed environments.

[0091] Step S200 includes at least steps S210-S230:

[0092] S210. Obtain the state initialization structure and construct the thermal-humid coupling state modeling structure.

[0093] This step is used to extract micro-environmental variables from the state initialization structure generated in S130, and to construct a state modeling structure that can express the coupled evolution of temperature and humidity. Combining the Navier-Stokes equations with an LSTM network, a time-differentiable thermo-humidity state tensor evolution system is constructed.

[0094] ① Construct the thermal and humidity input state tensor

[0095] The structure is initialized based on the state, and the current time t is extracted. k microenvironment temperature field T k (x,y,z), humidity field H k (x,y,z), condensation trend factor Γ k With the characteristic tensor Θ of the condensation region k Define the thermal and humid input state tensor S k as follows:

[0096]

[0097] in:

[0098] S k : The thermal and humidity input state tensor at the k-th time step;

[0099] T k (x,y,z): Current temperature distribution function;

[0100] H k (x,y,z): Current humidity distribution function;

[0101] Γ k : Condensation trend factor, a scalar that combines the condensation gradient and the slope of the local curve;

[0102] Θ k : The condensation region feature tensor represents the spatial morphology and density of the suspected condensation region;

[0103] d T d H d Θ : These represent the dimensions of temperature, humidity, and condensation structure, respectively.

[0104] ② Modeling the thermal and humidity evolution process based on the Navier-Stokes equations and LSTM:

[0105] Considering the nonlinear spatiotemporal diffusion characteristics of the temperature-humidity coupling process, the following composite evolution model is defined:

[0106]

[0107] in:

[0108] S k+1 : Represents the thermal-humidity coupling state structure at the (k+1)th time step. It is a state vector that reflects the coupling state of temperature, humidity and condensation factor in a certain area of ​​the chassis over time.

[0109] This represents a state prediction function based on a long short-term memory neural network (LSTM), used to model the time dependence and memory in thermal and humid state sequences;

[0110] Indicates the state S k The nonlinear correction function after correcting for the physical driving term and the disturbance term;

[0111] h k ,c k The hidden state and cell state of LSTM;

[0112]

[0113] This formula is the Navier-Stokes physical field constraint nonlinear adjustment function for the state, which describes the changing trend of temperature and humidity state in the continuous time domain under the combined effects of physical diffusion, pressure gradient and external disturbance.

[0114] in:

[0115] η: the diffusion regulation factor for warm and humid fluids;

[0116] Δ: Laplacian operator, used to express the diffusion term of the temperature and humidity tensor;

[0117] ΔS k : indicates S k The Laplace term, namely the spatial diffusion term of temperature and humidity, reflects the conduction and diffusion of heat and water vapor in space;

[0118] The pressure gradient term at the k-th time step corresponds to the driving force of mass migration due to pressure differences in heat and moisture transport.

[0119] f k External interference factors (such as cooling and current fluctuations);

[0120] The resulting thermo-humidity coupled state modeling structure possesses high-order temporal reasoning capabilities and incorporates fluid dynamics principles to predict the dynamic evolution trend of the temperature and humidity tensor over time.

[0121] S220. Based on the thermal-humidity coupled state modeling structure, calculate the state entropy increase sequence and humidity critical point.

[0122] This step is based on the thermo-humid evolution structure constructed in step S210. It uses a dual-channel entropy sequence modeling strategy that combines Shannon information entropy and thermodynamic entropy to calculate the microenvironment change rate and system reversibility index.

[0123] ③ Calculate Shannon entropy

[0124] Define a time window [t] k ,t k The probability P of the combined temperature and humidity distribution within +Δt] i,j Calculate the information entropy value of this window:

[0125]

[0126] in:

[0127] The combined temperature and humidity entropy value of the k-th window;

[0128] P i,j The probability density under the temperature and humidity combination distribution is generated based on the joint histogram of temperature and humidity.

[0129] log: Shannon entropy is calculated using the natural logarithm.

[0130] This formula, based on Shannon's information entropy theory, quantifies the statistical complexity and system order of the temperature and humidity state of the chassis within a certain time window. The higher the information entropy, the more dispersed and uncertain the temperature and humidity state distribution, and the more "chaotic" the microenvironment.

[0131] This formula provides a quantitative risk indicator for subsequent judgment of whether the system is close to condensation anomaly and whether active intervention is required;

[0132] Compared to the output value of a single sensor, it is better able to capture the changing trend of the system's "overall state fluctuations";

[0133] It provides key driving variables for entropy increase control algorithms, realizing a coupling bridge between environmental complexity and the intensity of regulation response.

[0134] ④ Calculate the entropy growth rate

[0135] Further define the entropy growth rate The rate of change of Shannon entropy per unit time:

[0136]

[0137] in:

[0138] Indicates the system from t k to t k+1 The entropy change trend represents the entropy growth rate within the k-th time period;

[0139] Δt: Time step (can be set to 5s or 10s);

[0140] Information entropy values ​​at consecutive time points.

[0141] This formula, with "the change of information entropy over time" as its core, assesses the speed at which a system's state evolves from order to disorder, and is a dynamic quantitative indicator of the system's "degree of instability".

[0142] The higher the entropy growth rate, the more severe the disturbance in the system within a short period of time, which may be a precursor to condensation. This indicator can be used as a "pre-response trigger" in the control mechanism to initiate control actions before the risk occurs. Compared with the absolute humidity value, this rate indicator has a stronger forward-looking and dynamic discrimination capability.

[0143] ⑤ Calculation of dynamic humidity critical point

[0144] Define the humidity critical point To determine the upper limit of humidity at the critical point of system reversibility, and combining the entropy growth rate with the current ambient temperature, the following function is constructed:

[0145]

[0146] in:

[0147] The dynamic humidity critical point at time k;

[0148] γ: Maximum allowable humidity threshold, usually set to 100;

[0149] Entropy growth rate;

[0150] The average value of the current temperature tensor (representing the background thermal intensity);

[0151] 1-tanh(·): This forms a nonlinear compression structure for the upper limit of humidity, enabling the system to automatically reduce the humidity threshold under severe disturbances.

[0152] This formula constructs an adaptive humidity safety threshold generation mechanism that couples the entropy growth rate with the current temperature background to dynamically adjust the humidity critical point for condensation judgment.

[0153] It implements an intelligent response mechanism that allows for lower tolerance humidity in more unstable environments; it ensures that the system can intervene in advance and actively control humidity in areas of high thermal disturbance to reduce the probability of condensation; it replaces traditional static humidity settings and provides high scene adaptability; and it improves the anti-condensation performance and operational stability of the whole system in multiple environmental scenarios.

[0154] S230. Normalize the state entropy increase sequence and the humidity critical point to generate a set of control mapping parameters.

[0155] This step standardizes the core indicators generated in the first two steps and outputs a control vector structure that is adapted to the subsequent control system.

[0156] ⑥ Control mapping parameter generation function

[0157] Define the control mapping parameter set as follows:

[0158]

[0159] in:

[0160] C k : This is the set of control mapping parameters, a three-dimensional vector representing the amount of control command execution;

[0161] Heating element power;

[0162] Dehumidifier fan speed;

[0163] Cooler current;

[0164] Control parameter mapping function;

[0165] k: Time window number;

[0166] This formula constructs a mathematical path that maps execution-level control parameters from the state entropy increase index and the humidity safety limit, realizing a quantitative transition from the state modeling layer to the control execution layer.

[0167] Mapping function The definition is as follows:

[0168]

[0169] in:

[0170] α i Entropy increase weight;

[0171] β i Temperature and humidity margin weighting;

[0172] Average humidity and temperature;

[0173] T target Target constant temperature setpoint.

[0174] When this formula is implemented, the cooling chip will increase its operating current when there is severe environmental disturbance or the temperature is higher than the target value; it will accurately pull back the temperature curve and effectively maintain thermal steady state; and it will avoid system balance drift caused by the accumulation of thermal disturbance.

[0175] The thermal-humid coupling state modeling structure and control mapping parameter generation mechanism constructed through steps S210–S230 have the following technical effects:

[0176] This invention achieves a dynamic coupled representation for multivariate thermal and humidity state modeling. It employs a modeling approach combining the Navier-Stokes fluid model and an LSTM time-recurrent neural network, effectively describing the temporal coupled evolution of temperature, humidity, and condensation trend factors. This overcomes the problems of weak univariate prediction capabilities and slow response to microenvironmental changes in traditional methods.

[0177] By introducing information entropy sequences to characterize the changing trends of the microenvironment, the system's self-sensing capabilities are enhanced. By defining state entropy increase sequences as an indicator of the degree of environmental order, the system can dynamically monitor environmental stability and predict potential mutation risks. This achieves the ability to uncover potential anomalies from quantitative fluctuations, improving the accuracy of judging condensation precursor states.

[0178] The humidity threshold is dynamically generated based on the entropy growth rate, improving the sensitivity of condensation detection and the ability to control it individually. The humidity threshold is generated by a nonlinear function of the entropy growth rate and the current temperature. The system can dynamically adjust the allowable humidity threshold under different thermal disturbance backgrounds, significantly improving the adaptability and robustness of the condensation control strategy.

[0179] The control mapping parameter set enables precise regulation of heating, dehumidification, and cooling. Based on the normalized entropy growth rate and humidity critical point, the control mapping parameter set can accurately output multi-dimensional regulation commands for heating power, fan speed, and cooling current, providing a high-precision parameter basis for subsequent power allocation and resource scheduling, and realizing a complete closed loop from perception modeling to precise execution.

[0180] This ensures the convergence and stability of the subsequent control command system. The set of control mapping parameters generated by this module is fully constrained in terms of spatial continuity and temporal evolution trend, ensuring the convergence and safety of the executed control command set (S330) under actual operating conditions, and avoiding control overshoot or instability caused by parameter mutations.

[0181] It promotes multi-stage linkage and closed-loop feedback updates of parameters in the overall system. The set of control mapping parameters not only provides the basis for execution parameters in S300–S330, but also serves as a benchmark for feedback updates in the S600 stage, ensuring that the state modeling structure can be accurately corrected after execution errors, thus forming a dynamic closed loop for thermal and humidity control.

[0182] Step S300 includes at least steps S310-S330:

[0183] S310. Obtain the set of control mapping parameters and construct the power allocation strategy and resource scheduling structure.

[0184] After generating the control mapping parameter set in step S230, the system enters the power regulation logic construction stage. In this step, the system first obtains the control mapping parameter set output by step S230. The control mapping parameter set has been integrated with the normalized entropy growth rate and humidity critical point parameters to characterize the temperature and humidity fluctuation trend and condensation critical risk in the current chassis microenvironment.

[0185] Specifically, the system constructs a power allocation strategy structure corresponding to the heating unit, cooling unit, and dehumidification unit based on the control fields in the control mapping parameter set. This strategy structure uses a multi-dimensional state decision vector and a resource supply capacity matrix for matching and fusion to ensure that each control unit executes the optimal control response under the premise of limited energy consumption.

[0186] Furthermore, based on the aforementioned power allocation strategy structure, the system combines the current power budget ceiling, the response delay coefficient of the thermosensitive material, and the start-up inertia data of the actuators to construct a resource scheduling structure. This resource scheduling structure includes a scheduling table for the start-up sequence, duration, and peak response capability of the three types of core actuators (heating element, dehumidifying fan, and cooling element) under different time periods and different condensation risk levels.

[0187] During the construction of the scheduling structure, the system introduces a condensation trend factor in the state initialization structure to enhance the priority control weight allocation of potential local condensation areas, thereby realizing a dynamic regional weight adjustment mechanism for resource scheduling.

[0188] S320: Perform scheduling calculations on the power allocation strategy and resource scheduling structure to obtain data on heating element power, dehumidifying fan speed, and cooling element current.

[0189] After completing the construction of the power allocation strategy and resource scheduling structure, the system enters the control parameter generation stage. Based on the aforementioned strategy and scheduling structure, the system performs multi-objective scheduling calculations and outputs three core control parameters: heating element power, dehumidifying fan speed, and cooling element current.

[0190] Specifically, the system first analyzes the trend of the current state entropy growth rate under the time-slice division strategy to determine whether there is a high risk of sudden change. If the judgment is correct, the response priority and speed threshold of the dehumidifying fan are appropriately increased; if it is determined to be a low-entropy fluctuation, the current of the cooling chip is reduced to save energy. This calculation process integrates the differential perturbation approximation method and the point-by-point dynamic adjustment mechanism to ensure that the control parameters are adjustable in real time within each sampling period.

[0191] For the distribution of heating element power, the system further introduces the local thermal coupling response factor in the structural characteristics of the condensation region, calculates and merges the required thermal compensation intensity for different regions, outputs the total heating power value, and performs local amplification adjustment according to the actual regional circuit distribution.

[0192] The dehumidifying fan speed is linearly modulated based on the deviation between the absolute value of the humidity critical point and the relative humidity of the current microenvironment, and a wind resistance load feedback parameter is added to the strategy to achieve load adaptive control of the fan operation.

[0193] The regulation of the thermoelectric current is based on the power budget curve and the trend of micro-environment temperature change. Through the sliding window integration mechanism, an adaptive current function range is output as the basis for the generation of subsequent control commands.

[0194] S330 integrates and processes the heating element power, dehumidifying fan speed, and cooling element current data to generate a set of execution control commands.

[0195] After acquiring the three core parameters—heating element power, dehumidifying fan speed, and cooling element current—the system enters the instruction fusion and control set construction stage. In this step, the system performs data alignment processing on the aforementioned three types of control parameters to construct a control parameter fusion structure with a unified time index and control cycle number.

[0196] Specifically, the system first uses the time base in the state initialization structure to unify the timestamp information of the control parameters and performs linear interpolation to complete the missing time periods. Then, by constructing a three-dimensional control command mapping matrix, the heating element power, dehumidifying fan speed, and cooling element current parameters are mapped to the corresponding execution interface address areas, and a set of control commands conforming to the protocol specifications is generated.

[0197] Furthermore, to improve the real-time performance and anti-interference capability of control execution, the system adds redundant verification fields and execution confirmation mechanisms to each instruction fragment in the control instruction set to ensure reliable execution of the control process in a high electromagnetic interference environment.

[0198] The system finally outputs the set of execution control instructions, and calls the set in step S510 to jointly control the starting sequence, running duration and dynamic response characteristics of the heating element, dehumidifying fan and cooling element, so as to complete the dynamic response regulation of the microenvironment.

[0199] Through the power allocation strategy construction, resource scheduling structure generation, and control command fusion mechanism described in steps S310–S330, a closed-loop linkage from state modeling to control execution is achieved, resulting in the following technical effects:

[0200] It can realize multi-dimensional control fusion across units and parameter types, ensuring the accuracy and consistency of the control response;

[0201] Based on the synergistic mapping mechanism between entropy increase sequence and condensation trend factor, priority thermal regulation resource allocation is realized in high-risk condensation areas to improve the anti-condensation effect;

[0202] The control command set has a unified format, protocol redundancy, and dynamic response as triple guarantees, enhancing the system's robustness and anti-interference capabilities;

[0203] The system has controllable power consumption and high execution efficiency during overall operation, and is suitable for chassis system application scenarios of different sizes and enclosure levels;

[0204] The control parameter structure and feedback structure have a high degree of coupling, providing a stable input basis for error analysis and parameter correction in subsequent S500–S600.

[0205] Step S400 includes at least steps S410-S430:

[0206] S410. Obtain the humidity critical point and condensation trend factor to determine the conditions for condensate formation.

[0207] This step aims to further call the historical data and current values ​​of the humidity critical point and condensation trend factor based on the control mapping parameter set generated in step S230, in order to determine whether the dynamic conditions for condensation generation have been met inside the chassis.

[0208] Specifically, the system first obtains the humidity critical point derived from the thermo-humidity coupled state modeling structure in step S220. This humidity critical point is a nonlinear time function characterizing the critical threshold for condensation formation. Its input comes from the microenvironmental data in the state initialization structure generated in S130, covering indicators such as temperature gradient changes, local humidity disturbance intensity, and heat flux density in the condensation region. Before executing this step, the system needs to synchronously call the set of control instructions generated in S330, which has already been executed within the current time slice, to ensure that the comparative analysis has real-time performance and execution synchronization.

[0209] The system extracts the condensation trend factor constructed in step S120. The condensation trend factor is a predictive index generated by the coupled evolution of the temperature and humidity change rate curve, the equivalent thermal resistance of the condensation region surface, and the humidity gradient. The system compares the current condensation trend factor value with the critical evolution surface model to determine whether it crosses the predicted extreme value domain.

[0210] The system will fuse and compare the humidity critical point with the condensation trend factor to determine whether the current state meets one of the following conditions: (1) The current humidity value is higher than the humidity critical point and the condensation trend factor is in a significantly increasing range; (2) The condensation trend factor is in a reverse growth state, but the humidity value is in a stable range near the upper limit of the critical point; (3) The system detects that the heat flux density change in the condensation area has undergone a double jump, and confirms the presence of condensation signs by combining the humidity critical point threshold.

[0211] When any condition is met, the system marks the current state as "Condensation generation conditions are met" and writes it into the condensation determination structure, which is used as the basis for logical judgment in the subsequent hydrophilic guide channel control and re-evaporation path configuration process.

[0212] S420. Control the state of the hydrophilic guide channel according to the condensate generation conditions.

[0213] After determining in step S410 that the conditions for condensate generation inside the chassis have been met, the system immediately enters the hydrophilic guide channel state control process. The hydrophilic guide channel, as the core structural component of the condensate self-recovery design of this invention, has a surface coated with a multi-layered nano-coating with high hydrophilicity, enabling fluid guidance even in the early stages of trace condensate generation.

[0214] Specifically, based on the condensate generation status marker in the condensation determination structure in step S410, the system sends an opening command for the hydrophilic guide channel. This opening command controls the micro-drive actuator of the guide channel to rotate and expand to a preset angle range, exposing the hydrophilic guiding surface. The expansion range of this angle needs to be matched with the cooling chip current data in step S320 to ensure the formation of an optimal liquid guiding path under the current heat flow channel conditions.

[0215] The system needs to verify the current opening status of the evaporation membrane. If the evaporation membrane is closed, a preheating and soft-start control command for the evaporation membrane will be sent before opening the guide channel to ensure that the condensate can flow smoothly into the evaporation area after entering the guide channel. If the guide channel is already open and the condensate has not yet formed a stable flow path, the system will maintain the current state of the guide channel and continuously collect feedback data on the guiding effect for subsequent response evaluation and model updates in S530.

[0216] During the process of controlling the hydrophilic guide channel, the system needs to periodically detect the environmental data and the change curve of the condensation trend factor. If an external environmental disturbance (such as a sudden drop in temperature or abnormal operation of the fan) is detected, causing the condensation trend to reverse, the system can send a guide channel closing command and close the guide path to prevent the guide path from idling and causing energy waste.

[0217] S430. Based on the state of the hydrophilic guide channel and the opening state of the evaporation membrane, establish a re-evaporation path configuration structure.

[0218] After completing the status control of the guide channel, the system will enter the re-evaporation path configuration process to realize the dynamic recovery and reuse of condensate resources.

[0219] Specifically, the system first acquires the physical open status of the hydrophilic guide channel and the heating status of the evaporation membrane, and then reads the latest data on the cooling chip workload and heating chip power from the integrated execution control command set in the S330. Based on this data, it determines the heat flux balance of each heat exchange unit inside the chassis and performs path matching accordingly.

[0220] When the hydrophilic guide channel is open, the evaporation membrane has been preheated, and the power of the internal heat exchange plates is in a stable range, the system establishes a re-evaporation path configuration structure. This structure contains the following fields: (1) current condensate flow rate range; (2) guide channel conduction period identifier; (3) evaporation membrane target temperature zone configuration parameters; (4) evaporation return node connection path; and (5) matching parameters with the thermal channel configuration of the refrigeration area.

[0221] To improve the robustness and cycle efficiency of the path configuration, the system will combine the state entropy increase sequence generated by S220 to judge the heat transfer stability of the re-evaporation path under the current environment. If the system entropy increase rate is in a stable or decreasing range, the multi-segment path parallel configuration strategy can be enabled to achieve concurrent recycling of multi-source condensate. Otherwise, if the state entropy increase is in an oscillating state, a flow-limiting path structure will be adopted, retaining only the single-channel return logic to prevent the risk of thermal disturbance feedback coupling.

[0222] Once the re-evaporation path configuration structure is established, it will be written into the control structure interface in S510 and used as the path configuration benchmark in the feedback acquisition structure for subsequent microenvironment response evaluation and system adaptive model update.

[0223] By executing steps S410–S430 in this embodiment of the invention, the system achieves prediction and determination of condensate generation conditions inside the sealed enclosure, precise control of the hydrophilic guide channel state, and dynamic configuration of the re-evaporation path, resulting in the following technical effects:

[0224] The system can pre-identify the risk of condensation before condensate is actually generated by using a dual judgment mechanism of humidity critical point and condensation trend factor, thus avoiding short circuits or corrosion of electronic components caused by delayed response.

[0225] The opening and closing process of the hydrophilic guide channel is based on intelligent prediction and real-time feedback closed loop, no longer relying on a fixed timing structure, which significantly improves the response capability to complex environmental changes.

[0226] The generation of the re-evaporation path configuration structure integrates the current thermal and humidity state with the dynamic configuration parameters of the condensation structure, ensuring that the condensate can be quickly evaporated and recovered after being guided, forming a true closed loop of moisture and heat coupling control.

[0227] This step forms a closed-loop parameter calling relationship with steps S100–S300, and provides a complete structure and state basis for feedback collection and model updates in steps S500–S600, thereby improving the overall system's adaptability and coordination.

[0228] The deep integration of structural physical control (flow channel) with dynamic modeling algorithms (entropy increase and trend factor) has strong interdisciplinary collaborative design value and is a key supporting path for realizing the innovation of "condensate self-recovery design".

[0229] Step 500 includes at least steps S510-S530:

[0230] S510: Obtain the set of execution control instructions and the re-evaporation path configuration structure, and jointly control the working status of each execution unit.

[0231] In step S510, the system first reads the execution control instruction set from S330 and the re-evaporation path configuration structure from S430. The execution control instruction set includes multi-dimensional control parameters such as heating element power command, dehumidifying fan speed command, and cooling element current command. This set has been generated based on the control mapping parameter set and has execution properties that match the current chassis state. The re-evaporation path configuration structure originates from the evaluation of the hydrophilic guide channel state and evaporation membrane opening state in S430, and includes multiple fields such as path identifier, number of channels, time period, recirculation node, and structural node.

[0232] The system combines the two to synchronize the status of the control execution units. Specifically, the system sends the set of execution control commands and the re-evaporation path configuration structure to the execution response module (corresponding to the execution response module 60 and the execution feedback acquisition and parameter update module in the system structure diagram), which triggers the coordinated operation of hardware units including heating elements, dehumidifying fans, semiconductor cooling chips, and hydrophilic guide channel actuators. The system internally generates multi-channel bus control logic synchronously to ensure synchronized command execution.

[0233] Furthermore, the system activates the execution unit status monitoring mechanism to collect real-time status variables such as operating voltage, current, rotational speed, and mechanism position of each execution unit, and writes the collected results into a temporary status structure. This temporary status structure, together with the re-evaporation path configuration structure, is used for subsequent response monitoring and feedback analysis.

[0234] When the system is under joint control, it relies on the set of control mapping parameters passed from steps S300–S330 as the basis for correction and performs “closed-loop calibration” at the hardware control level. For example, if the initial starting speed of the dehumidifier fan is found to be too low, the system can superimpose several preset pulse control signals to ensure that it quickly enters the target speed range; if the evaporator film heating has not yet reached the specified temperature, the system will synchronously adjust the power of the heating element.

[0235] This implementation step ensures that the control strategy obtained by the preceding modules through condensation trend analysis (S130), thermal-humidity coupling modeling (S210–S230), and power scheduling (S310–S330) is output to the actual hardware execution system, completing the closed-loop connection between the theoretical and physical levels.

[0236] S520: During the execution of control commands, acquire temperature and humidity change data inside the chassis and generate a microenvironment response sequence.

[0237] After the execution control process enters the implementation phase, the system enters step S520, which is used to monitor the control feedback status and generate response data.

[0238] Specifically, after the execution unit starts working, the system collects temperature and humidity values ​​inside the chassis through multi-point temperature and humidity sensors in the environmental data acquisition module 10, and collects status information of the condensation area through the condensation monitoring device. The temperature and humidity change data are in time series format, which facilitates the construction of subsequent response sequences.

[0239] The system sets the sampling interval of the collected data to be synchronized with the control command output cycle (e.g., once per second or once per half second, depending on the control mapping parameter set), automatically records it in the response timestamp, and supplements the sampled values ​​one by one into the microenvironment response sequence. The microenvironment response sequence structure includes a timestamp, temperature value, humidity value, control command identifier code, and current path status field. This sequence originates from the immediate feedback after hardware execution in step S510 and possesses complete response mapping properties.

[0240] During the generation of the response sequence, the system compares the current response data with the humidity critical point generated in S220 and the condensate generation conditions evaluated in step S410 in real time. When the response data changes drastically (such as the temperature dropping instantly by more than ±0.5℃ or the humidity suddenly changing by close to 40%), the system immediately marks the corresponding response record as "high disturbance response" and writes it into the response structure.

[0241] The entire S520 process establishes a response perception closed loop for the system, forming a clear one-to-one feedback path with the control commands output by the preceding S300, providing a foundation for subsequent error analysis.

[0242] S530: Perform dynamic error calculation and control response analysis on the microenvironment response sequence to generate a feedback acquisition structure.

[0243] After the response data collection is completed in S520, the system enters S530 to perform dynamic error analysis on the response results and generate a feedback structure.

[0244] The system associates the response effect of each control command with the timestamps of the executed control command set and the microenvironment response sequence, and compares the difference between the actual temperature and humidity changes and the target temperature and humidity changes to form a basic error sequence. The error sequence is a dynamic numerical structure that includes attributes such as error amplitude, error duration, error rate of change, and error direction indicator.

[0245] Specifically, the system generates multiple statistical indicators, such as:

[0246] Maximum temperature error;

[0247] Average humidity error;

[0248] Error duration;

[0249] Error information fluctuation range.

[0250] Furthermore, the system performs trend analysis and control response efficiency evaluation on the error sequence through the response analysis module, maps the error judgment results to the control response quality index, and evaluates each execution unit (heating element power response time, dehumidifying fan speed response delay, cooling element cooling efficiency, hydrophilic guide channel flow path stability, etc.).

[0251] The system packages the aforementioned control response quality indicators and error information into a feedback acquisition structure, which includes the following fields:

[0252] Control command ID;

[0253] The error between the target and the actual response;

[0254] Response time;

[0255] Execution unit running status pointer;

[0256] Path configuration change identifier;

[0257] Rate of change of thermal-humid coupling state.

[0258] The system provides the feedback acquisition structure to the S600 as a basis for model correction and parameter updates. Through the feedback acquisition structure, the system can identify deficiencies in the control mapping parameter set. For example, if a certain control parameter fails to reduce humidity rapidly after actual execution, the feedback structure will record this state for the S600 to use in correcting the thermal-humidity coupling state modeling structure and power scheduling parameters.

[0259] By implementing steps S510–S530, the system forms a closed-loop execution mechanism from control implementation to real-time response and error feedback, achieving the following technical effects:

[0260] Seamless integration of control execution and response perception is achieved. The joint control of the execution control command set and the re-evaporation path configuration structure ensures that the state of each execution unit is synchronized with the control strategy, thereby improving the timeliness and synchronization of control response.

[0261] A high-precision microenvironment response sequence is constructed. By sampling temperature and humidity changes at high frequency and generating response sequences, a fine characterization of the microenvironment state inside the chassis is achieved, providing high-quality data support for error calculation.

[0262] It provides a feedback structure based on dynamic error and response analysis. The feedback acquisition structure transforms the response error into a basis for control adjustment and has a regulation efficiency index, which is used to drive the system to adaptively update parameters.

[0263] To ensure the system's closed-loop adaptive control capability, this module serves as the basis for S600 parameter updates, enabling the preceding mapping parameters and modeling structure to be adaptively corrected based on the feedback acquisition structure, thereby enhancing system stability and long-term operating performance.

[0264] The overall efficiency of system collaborative control is improved. The S510–S530, S210–S230, and S310–S330 modules form a vertical closed-loop structure, integrating environmental perception, modeling, control execution, and model correction into one, realizing a control system with strong dynamic adaptability and multi-execution unit collaboration.

[0265] Step S600 includes at least steps S610-S630:

[0266] S610. Obtain the set of feedback acquisition structure and control mapping parameters, and construct the state mapping and control relationship table.

[0267] In step S610, the system first obtains the feedback acquisition structure from step S530. This structure contains key data such as error information between the executed control command and the actual temperature and humidity response, control execution efficiency indicators, execution delay, and path configuration change identifiers. Simultaneously, it obtains the control mapping parameter set from step S230, including heating element power control mapping parameters, dehumidifier fan speed control mapping parameters, and cooling element current control mapping parameters, along with corresponding parameter sequences and normalization coefficients.

[0268] Understandably, the system further uses the two datasets as input to construct a "state mapping and control relationship table". Specifically, this relationship table is designed as a two-dimensional mapping structure. The horizontal index represents the error and response index fields of the time series in the feedback acquisition structure (such as maximum temperature error, average humidity error, response time delay, etc.), and the vertical index represents the control variable parameters in the control mapping parameter set. The content of this relationship table is stored in key-value pair format, where the key is the feedback state field and the value is the corresponding control variable weight adjustment suggestion.

[0269] Specifically, the system uses a feature mapping algorithm to associate the dynamic temperature and humidity error with the current control mapping parameters and calculates the adjustment weight of each control parameter under the current error state. For example, when the feedback acquisition structure indicates that the humidity response lag continues to exceed a preset threshold, the state mapping and control relationship table will correspondingly increase the weight of the dehumidifier fan speed and the heating element power correction suggestion.

[0270] The construction process also includes a data stability detection mechanism to identify noise and fault data; and automatically refits the mapping weights using historical mapping data to ensure that the relational table constructed each time matches the current control environment and deviation.

[0271] S620. Based on the state mapping and control relationship table, correct the thermal-humid coupling state modeling structure and power allocation strategy parameters.

[0272] In step S620, the system uses the suggestions in the state mapping and control relationship table to perform correction operations on the thermal-humid coupling state modeling structure and power allocation strategy parameters respectively.

[0273] Specifically, for the thermal-humidity coupled state modeling structure, the coupling function model and entropy increase sequence generation logic constructed in the preceding S210 have already been loaded into this structure. The system evaluates the fitting accuracy of the current modeling structure by checking the "state change rate" and "path configuration change indicator" in the feedback acquisition structure. If a persistent deviation in the average temperature and humidity error is found, or if the fitted thermal-humidity coupling curve deviates from the actual response curve beyond the predetermined error tolerance, the system automatically adjusts the coupling coefficient and hysteresis coefficient in the modeling module to make the thermal-humidity modeling structure more closely match the current execution feedback capability.

[0274] The correction operations encompass the reweighting of the temperature gradient term in the entropy increase function and the adjustment of the response compensation for the humidity delay term, to enable the structure to adapt to the current environmental state. Simultaneously, the system can fine-tune the entropy increase judgment threshold (humidity critical point) calibration strategy, for example, by adjusting the threshold sensitivity based on the frequency of condensation response from the feedback-collected structure, thereby improving the adaptability of the established state-modeled structure.

[0275] The system modifies the power allocation strategy parameters. These parameters include the heating element power allocation priority, the cooling element current allocation priority, and the dehumidifier fan speed allocation ratio. The system updates the parameter values ​​by analyzing the control weight suggestions in the state mapping and control relationship table. For example, if the feedback structure indicates insufficient humidity response, the system increases the dehumidifier fan speed weight and correspondingly decreases the heating element power priority; if the path configuration shows lag in the guide channel pipeline, the cooling element current adjustment range is adjusted accordingly.

[0276] The parameter corrections are implemented incrementally or incrementally, avoiding a significant one-time change to the overall control strategy to ensure system stability. The corrected parameters are saved as a new power allocation parameter set for use in subsequent control loops to output more precise control commands.

[0277] S630. Update the corrected thermal-humid coupling state modeling structure and power allocation strategy parameters to the state initialization structure.

[0278] In step S630, the system carries the new thermal-humid coupling state modeling structure and the revised power allocation strategy parameter set. The system encapsulates these two elements back into the state initialization structure, replacing the original structure, so that the closed-loop operations of S200, S300-S530 can be re-executed in the next control cycle.

[0279] Specifically, the system will update the following fields in the state initialization structure:

[0280] Thermal-humid coupling model structural identifier: indicates the use of a new structure or the inclusion of correction records;

[0281] Entropy increase sequence calculation function pointer: points to the calibrated new version of the entropy increase algorithm;

[0282] Humidity critical point calculation module update flag: Preference value has changed or judgment logic has been modified;

[0283] Power allocation strategy parameters: updated priority coefficient, normalization coefficient, adjustment range, etc.

[0284] Understandably, to ensure execution continuity, the update operation also includes version number management, timestamp recording, and a structure consistency verification mechanism. After the structure update is completed, the system automatically switches to the next control cycle, where S210 again models and calculates entropy increase for the new state initialization structure, forming a fully closed-loop logic.

[0285] By introducing a "state mapping and control relationship table" based on the feedback acquisition structure and control mapping parameter set in steps S610–S630, this invention achieves an adaptive closed-loop update mechanism for the regulation model and control strategy. This mechanism has the following technical effects:

[0286] This invention achieves dynamic correction of the control strategy based on multi-source feedback information. Compared to traditional systems based on fixed parameter control, this invention introduces a state mapping relationship construction method for the first time in the temperature and humidity control of high-sealing electronic enclosures. This allows the system to accurately adjust the control mapping parameters based on feedback information such as temperature and humidity errors, response lag, and changes in the condensation path. This, in turn, enables dynamic re-optimization of the control commands, resulting in a more agile and precise coordinated response from the heating element, dehumidifying fan, and cooling element.

[0287] To improve the adaptability and fitting accuracy of thermal and humidity modeling structures to changes in the microenvironment, the original modeling structure may exhibit modeling mismatch or response lag after long-term operation in a closed environment. This invention identifies fitting deviations in the modeling structure through state mapping relationships and automatically corrects the thermal and humidity coupling function, hysteresis parameters, and entropy increase sequence generation logic, thereby ensuring that the system's modeling logic dynamically adjusts with the environment and remains distortion-free during long-term operation.

[0288] A continuously evolving control system structure is constructed. By re-encapsulating the modified modeling structure and control parameters into a state initialization structure in step S630, the structure version is updated iteratively. This allows the system to directly use the optimal modeling parameters in the next cycle, significantly improving control efficiency, avoiding environmental reset and parameter failure, and constructing a truly "closed-loop" control framework.

[0289] To enhance the robustness and anti-condensation stability of the system, the feedback-driven control and update mechanism constructed in this invention enables the system to handle high humidity and high heat coupled disturbances. Especially in the face of abnormal situations such as sudden environmental changes and a surge in chassis heat sources, it can still complete strategy calibration and model reconstruction within a limited response time, effectively avoiding the sudden generation of condensate and improving the stability of the internal environment of the chassis and the reliability of equipment operation.

[0290] It supports the evolution of low-power and fine-grained control strategies. The system adjusts the priority of resource allocation through feedback error, realizes on-demand power readjustment of the three paths of thermal control, dehumidification and current, reduces over-control and resource redundancy, improves energy efficiency, and extends the life of the fan and heating module, with good energy-saving characteristics.

[0291] The key innovations of this invention include:

[0292] (1) Modeling structure of thermal and humid state driven by condensation trend factor and entropy increase sequence. The characteristics of condensation area are dynamically integrated with temperature and humidity gradient, and an entropy increase sequence is defined to reflect the change in environmental orderliness, forming a modeling foundation that is more predictive and generalizable than traditional temperature and humidity thresholds.

[0293] (2) Multi-execution unit joint control mechanism driven by control mapping parameter set. A control mapping parameter structure with normalized entropy growth rate and humidity critical point as the core is constructed to support the linkage control of heating, dehumidification and refrigeration equipment, and has the advantages of high control accuracy, flexible resource allocation and stable control convergence.

[0294] (3) Adaptive update mechanism linking re-evaporation path configuration structure and feedback acquisition structure. By performing error calculation and strategy inversion on the execution control results and microenvironment response sequence, the state modeling structure and control strategy parameters are updated in real time, ensuring the robustness and convergence of the system under various operating conditions.

[0295] The following are its main beneficial effects:

[0296] (1) Improve the response accuracy and predictive foresight of microenvironment perception. This invention overcomes the problems of delayed response and slow regulation of traditional systems to condensation generation state by introducing a joint modeling approach of "condensation trend factor" and "entropy increase sequence". The system uses time series features extracted from temperature and humidity data to dynamically calculate the rate of increase of state entropy, and combines the Navier-Stokes fluid model with thermal expansion.

[0297] (2) Achieving active and adjustable heat and humidity control and multi-source resource coordination. This invention innovatively constructs a "control mapping parameter set" based on normalized entropy increment and humidity critical point. This set is used to simultaneously drive the coordinated work of multiple execution units such as heating element power, dehumidifying fan speed and semiconductor cooling chip current. Compared with traditional temperature control systems that can only respond at a single point, this system can form a multi-dimensional control strategy combination according to different environmental disturbances, ensuring that the system can maintain a highly stable operating environment with temperature ±1℃ and humidity ≤40% under complex background of heat and humidity coupling.

[0298] (3) Construct a dynamic feedback-driven state modeling correction mechanism to enhance long-term adaptability and system robustness. During the execution of the control strategy, the system introduces a "micro-environment response sequence" and a "feedback acquisition structure" to dynamically calculate the error and backtrack the deviation between the actual temperature and humidity changes and the predicted state after execution. Then, through error weight remapping, the thermal-humidity coupling state modeling structure and control parameters are automatically corrected, which reduces the risk of failure caused by model drift and hardware aging during long-term system operation while maintaining control accuracy, thus forming a closed-loop control capability with "self-sensing - self-adjustment - self-adaptation".

[0299] Example 2: Figure 2 A structural block diagram of a constant temperature chassis environment sensing and intelligent control system according to an embodiment of the present invention is shown. Figure 2 As shown, the structure may include:

[0300] The environmental data acquisition module 10 is used to acquire temperature, humidity, and condensation area data inside the chassis. It uses multi-point distributed temperature and humidity sensors and condensation monitoring devices to achieve real-time sensing of micro-environment information. This module has outlier removal and format standardization functions, generating standardized micro-environment data from the raw environmental data for subsequent condensation trend calculation and state initialization.

[0301] The condensation trend analysis and state initialization module 20 is used to extract temperature and humidity change curves, local humidity gradients, and condensation region characteristics from microenvironment data. By calculating the condensation trend factor and fusing it with the microenvironment data, a state initialization structure is generated. This structure serves as input for thermal and humidity modeling, containing initial state information and condensation risk assessment indicators.

[0302] The modeling and entropy increase calculation module 30 is used to construct a thermal-humidity coupled state modeling structure and calculate the state entropy increase sequence and humidity critical point based on this structure. Internally, the module uses coupled modeling functions to express the temperature-humidity interaction relationship and, combined with a dynamic evolution mechanism, outputs the current entropy increase rate and the humidity threshold for condensation determination. After normalization, a set of control mapping parameters is generated as the basis for the regulation strategy.

[0303] The control strategy generation and instruction construction module 40 is used to construct power allocation strategies and resource scheduling structures based on the control mapping parameter set, and to perform scheduling calculations accordingly. By jointly analyzing the current entropy state and critical humidity value, it outputs key control parameters such as heating element power, dehumidifying fan speed, and cooling element current. Finally, the control parameters are integrated into a unified set of execution control instructions for collaborative use by multiple execution units.

[0304] The condensate recovery control module 50 is used to determine the condensate generation conditions based on the humidity critical point and condensation trend factor, and control the state of the hydrophilic guide channel accordingly. When the generation conditions are met, the system opens the hydrophilic guide channel to guide the condensate to the evaporation film area. The module further establishes a re-evaporation path configuration structure based on the state of the guide channel and the opening state of the evaporation film, supporting the management of condensate recovery and circulating evaporation path.

[0305] The execution feedback acquisition and parameter update module 60 is used to acquire the set of execution control commands and the re-evaporation path configuration structure, jointly control the working status of the heating element, cooling element, and dehumidifying fan, and collect temperature and humidity changes inside the chassis during control execution to generate a microenvironment response sequence. The module performs error calculation and response analysis on the response data, generates a feedback acquisition structure, and, combined with the control mapping parameter set, constructs a state mapping and control relationship table, corrects the thermal-humidity coupling state modeling structure and power allocation strategy parameters, updates the state initialization structure, and forms a complete closed loop.

[0306] The constant-temperature chassis environmental sensing and intelligent control system provided in this embodiment achieves dynamic modeling and precise control of the internal temperature and humidity by constructing a thermo-humidity coupled state modeling structure with entropy minimization as its core, and integrating condensation trend analysis and humidity critical point prediction. The system incorporates a hydrophilic guide channel and evaporation film structural design to proactively respond before condensation occurs, intelligently triggering condensation recovery path configuration and effectively controlling condensation formation. Simultaneously, the system introduces a micro-environment response feedback mechanism to adaptively adjust the control strategy, forming a multi-source information-driven, fully closed-loop intelligent control system. This system boasts advantages such as high control accuracy, fast response speed, and high structural integration, making it suitable for environmental management scenarios of highly sealed, high-precision electronic equipment.

[0307] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. A constant-temperature machine case environment sensing intelligent regulation method, characterized in that, The method comprises the following steps: Obtain temperature and humidity data and condensation region data, extract a condensation trend factor, and fuse to generate a state initialization structure; the temperature and humidity data comprise instantaneous temperature values and humidity values of a plurality of sampling nodes and time sequence records; the condensation region data comprise local condensation state identification values collected by a plurality of condensation sensors; Obtain the state initialization structure, construct a heat and humidity coupling state modeling structure, calculate an entropy increase sequence and a humidity critical point, and generate a control mapping parameter set; The expression of the heat and humidity coupling state modeling structure is: wherein, is a thermal-hygro coupling state structure at the k+1th time step; is a state prediction function based on a long short-term memory neural network (LSTM); denotes a correction of the state by a physical driving term and a disturbance term; are the hidden state and the cell state of the LSTM; The expression of the control mapping parameter set is: in, To control the set of mapping parameters; This refers to the power of the heating element; This refers to the dehumidifier fan speed; This refers to the current of the thermoelectric cooler. This is a function for mapping control parameters; Number the time windows; This represents the rate of increase of entropy during the k-th time period; For the first Dynamic humidity threshold at all times; Based on the control mapping parameter set, construct a power distribution strategy and a resource scheduling structure, perform data integration processing, and generate an execution control instruction set; Determine a condensate generation condition according to the humidity critical point and the condensation trend factor, control a hydrophilic flow guide groove state, and establish a re-evaporation path configuration structure; Obtain the execution control instruction set and the re-evaporation path configuration structure, collect temperature and humidity changes, and generate a microenvironment response sequence, calculate an error, and generate a feedback collection structure after response analysis; specifically, obtain the execution control instruction set and the re-evaporation path configuration structure, jointly control the working states of a heating sheet, a refrigeration sheet, and a dehumidification fan; collect temperature change data and humidity change data in the control process, and generate a microenvironment response sequence; perform dynamic error calculation and response analysis on the microenvironment response sequence, and generate a feedback collection structure; According to the feedback collection structure and the control mapping parameter set, correct the heat and humidity coupling state modeling structure, and update to the state initialization structure; specifically, obtain the feedback collection structure and the control mapping parameter set, construct a state mapping and control relationship table; according to the state mapping and control relationship table, correct the heat and humidity coupling state modeling structure and the power distribution strategy parameters; update the corrected heat and humidity coupling state modeling structure and the power distribution strategy parameters to the state initialization structure.

2. The constant-temperature case environment-aware intelligent regulation method according to claim 1, characterized in that, The step of obtaining the temperature and humidity data and the condensation region data, extracting the condensation trend factor, and fusing to generate the state initialization structure comprises: Obtain temperature data, humidity data, and condensation region data in a case, perform outlier rejection and format unification, and obtain microenvironment data; Extract a temperature and humidity change curve, a local humidity gradient, and a condensation region feature from the microenvironment data, and calculate a condensation trend factor; Fuse the condensation trend factor and the microenvironment data to generate the state initialization structure.

3. The constant-temperature case environment-aware intelligent regulation method according to claim 1, characterized in that, The step of constructing the heat and humidity coupling state modeling structure comprises: Obtain the state initialization structure, and construct the heat and humidity coupling state modeling structure.

4. The constant-temperature case environment-aware intelligent regulation method according to claim 1, characterized in that, The step of calculating the entropy increase sequence and the humidity critical point comprises: Based on the heat and humidity coupling state modeling structure, construct an entropy increase sequence for representing a microenvironment change trend; Calculate a humidity critical point according to the entropy increase sequence, which is used to represent a condensation occurrence determination threshold.

5. The constant-temperature case environment-aware intelligent regulation method according to claim 1, characterized in that, The step of generating the control mapping parameter set comprises: Normalize the entropy increase sequence and the humidity critical point to generate a control mapping parameter set containing power, wind speed, and current control information.

6. The constant-temperature case environment-aware intelligent regulation method according to claim 1, characterized in that, The step of constructing the power distribution strategy and the resource scheduling structure comprises: Obtain the control mapping parameter set, construct the power distribution strategy and the resource scheduling structure, and The scheduling calculation is performed on the power distribution strategy and the resource scheduling structure to obtain heating sheet power, dehumidification fan rotating speed and refrigeration sheet current data; The heating sheet power, dehumidification fan rotating speed and refrigeration sheet current data are integrated and processed to generate an execution control instruction set.

7. The constant-temperature case environment-aware intelligent regulation method according to claim 1, characterized in that, The control of the hydrophilic flow guide groove state and the establishment of the re-evaporation path configuration structure body include: Obtaining humidity critical point and condensation trend factor to judge condensate generation condition; According to the condensate generation condition, the opening or closing state of the hydrophilic flow guide groove is controlled; According to the state of the hydrophilic flow guide groove and the opening state of the evaporation membrane, the re-evaporation path configuration structure body is established.

8. An intelligent control system for a constant-temperature machine case environment, applied to the intelligent control method for a constant-temperature machine case environment according to any one of claims 1-7, characterized in that, It includes: An environmental data acquisition module is used to obtain temperature and humidity data and condensation area data, and generate a state initialization structure body; The condensation trend analysis and state initialization module is used to extract temperature and humidity change curve, local humidity gradient and condensation area characteristics from micro-environment data; The modeling and entropy increase calculation module is used to build a heat and humidity coupled state modeling structure, and calculate the state entropy increase sequence and humidity critical point based on the structure; The regulation strategy generation and instruction construction module is used to construct the power distribution strategy and the resource scheduling structure based on the control mapping parameter set, and to perform scheduling calculation accordingly; The condensate water recovery control module is used to judge the condensate water generation condition according to the humidity critical point and the condensation trend factor, and to control the state of the hydrophilic flow guide groove accordingly; The execution feedback acquisition and parameter update module is used to obtain the execution control instruction set and the re-evaporation path configuration structure body, to jointly control the working state of the heating sheet, the refrigeration sheet and the dehumidification fan, and to collect the temperature and humidity changes in the cabinet during the control execution process to generate a micro-environment response sequence.

Images