Airflow and moisture coupling model construction and dynamic regulation method and system

Abstract

The present application relates to the technical field of mine ventilation and air conditioning, and discloses a wind flow heat and humidity coupling model construction and dynamic regulation method and system, which is suitable for temperature and humidity and thermal comfort control of key areas such as mining faces and electromechanical chambers in deep mines; the method comprises the following steps: collecting multi-source temperature and humidity sensing data and wind network topology, and generating a heat and humidity disturbance phase space grid; based on the grid, a heat and humidity compliance matrix is output by a heat and humidity conduction matrix generator; under the constraint of the matrix, a regulation strategy partial order lattice is generated by a multi-chamber conflict resolution partial order lattice optimizer; according to the partial order lattice, dynamic closed-loop regulation is performed by a multi-actuator guided by the lattice path; the present application solves the problem that traditional methods are difficult to multi-dimensionally couple the heat and humidity environment and the regulation is prone to conflict or overrun under the condition of multi-source disturbance and strong coupling in deep mines, realizes stable temperature and humidity of key areas in the compliance and comfort interval, and guarantees work safety and equipment stability.

Description

Technical Field

[0001] This invention relates to the field of mine thermal and humidity control technology, and more specifically, to a method and system for constructing and dynamically controlling a coupled airflow, thermal, and humidity model. Background Technology

[0002] In deep mine operations, mine thermal and humidity control technology is mainly used to address the coupled effects of surrounding rock heat exchange, equipment heat dissipation, and the superposition of concentrated heat sources on the temperature and humidity of mine airflow, maintaining stable temperature, humidity, and wet-bulb / black-bulb temperature index in key areas such as the mining face and pump room. This technology needs to be combined with the mine ventilation network topology, considering the transmission effect of periodic fluctuations in surface climate on intake air temperature. By constructing an airflow-thermal-humidity coupling model, it can predict the thermal and humidity status of key areas, distribute thermal and humidity loads, and implement dynamic closed-loop control to meet the compliant operation requirements of electromechanical equipment chambers and the operational comfort and efficiency needs of the mining face.

[0003] However, existing methods for controlling temperature and humidity in mines struggle to simultaneously adapt to multi-source disturbances, strong coupling characteristics, and rigid compliance constraints. This problem arises because the thermal and humidity environment in deep mines is deeply influenced by the coupled effects of surrounding rock heat exchange, equipment heat dissipation, and concentrated heat sources. Furthermore, the intake air temperature is affected by periodic fluctuations in surface climate, which are transmitted to various air-using areas through the ventilation network topology. Existing methods lack the dynamic response capability to such multi-source disturbances and do not fully integrate the transmission characteristics of the ventilation network topology with the rigid requirements of compliance constraints. This leads to temperature and humidity levels in critical areas easily exceeding specified thresholds, making it difficult to maintain stable comfort control targets in the operating area, thereby affecting the operational safety and efficiency of electromechanical equipment. Summary of the Invention

[0004] To address the technical challenges of traditional methods in depicting the thermal and humidity environment from multiple dimensions and easily leading to conflicts or exceeding limits in deep mines under multi-source disturbance and strong coupling scenarios, and to achieve stable temperature and humidity in key areas within a compliant and comfortable range, ensuring safe underground operations and stable operation of electromechanical equipment, this invention provides the following technical solution:

[0005] A method for constructing and dynamically controlling a coupled airflow-thermal-humidity model includes:

[0006] Step S10: Collect multi-source temperature and humidity sensor data and wind network topology, fuse the multi-source temperature and humidity sensor data and wind network topology, and generate and output the thermal and humidity disturbance phase space grid.

[0007] Step S20: Based on the thermal and moisture perturbation phase space grid, deduce the compliance evolution path through the thermal and moisture conduction matte generator and output the thermal and moisture compliance matte;

[0008] Step S30: Under the thermal and humidity compliance matrix constraint, generate a partial order lattice for the control strategy through a partial order lattice optimizer for multi-chamber conflict resolution;

[0009] Step S40: Based on the partial order grid of the control strategy, execute dynamic closed-loop control commands through multiple actuators guided by the grid path; the multiple actuators include damper actuators, chiller unit actuators and fan frequency converter actuators.

[0010] Furthermore, the fusion processing of multi-source temperature and humidity sensor data with the wind network topology includes:

[0011] Spatiotemporal alignment processing is performed on multi-source temperature and humidity sensor data to obtain spatiotemporal aligned data;

[0012] The spatiotemporal weights are calculated, which include time weights and spatial weights. The time weights are determined based on the time difference between the data acquisition time and the current time and the seasonal cycle variation. The spatial weights are determined based on the straight-line distance between the sensor and the spatial node and the wind resistance coefficient in the wind network topology.

[0013] The temperature, humidity, wind speed, and heat source intensity in the spatiotemporally aligned data are multiplied by their corresponding time weights and spatial weights to obtain weighted temperature data, weighted humidity data, weighted wind speed data, and weighted heat source intensity data.

[0014] Weighted temperature data, weighted humidity data, weighted wind speed data, and weighted heat source intensity data are used as quadruples of perturbation dimensions to form a thermal and humid perturbation phase space grid.

[0015] Furthermore, the thermal and humid disturbance phase space grid is a three-dimensional tensor structure, with the tensor dimensions being the spatial node dimension, the time step dimension, and the disturbance dimension. The spatial node dimension includes the locations of all key areas that need to be monitored within the mine, the time step length is divided based on the seasonal periodic fluctuations of the mine's surface climate, and the disturbance dimension elements are temperature, humidity, wind speed, and heat source intensity.

[0016] Furthermore, the thermal and moisture conduction matrix generator includes:

[0017] Extract the thermal and humidity state parameters of each spatial node at different time steps from the thermal and humidity disturbance phase space grid;

[0018] Establish rules for the transfer of thermal and moisture states based on thermal and moisture state parameters;

[0019] Starting from the initial thermal and humid state at the initial time step, a breadth-first search strategy is used to generate the future thermal and humid state evolution path;

[0020] Each generated thermal and humid state evolution path is verified for compliance to obtain a compliant evolution path;

[0021] Based on the compliance evolution path, a thermal and humidity compliance matte is constructed using a matte construction algorithm.

[0022] Furthermore, the matroid construction algorithm includes:

[0023] Define the base set as the set of all compliance evolution paths that have undergone compliance verification;

[0024] Define the family of independent sets as a subset of a basic set that satisfies the matroid axioms, which include the heritability axiom and the commutativity axiom;

[0025] The matroid basis is determined to be the independent set with the largest number of elements in the family of independent sets;

[0026] Iterate through all elements in the family of independent sets and verify whether they satisfy the axioms of inheritance and commutativity.

[0027] The structure of the thermo-humidity compliant matte consists of a base set, a family of independent sets, and a matte base.

[0028] Furthermore, the partial order lattice optimizer for multi-chamber conflict resolution includes:

[0029] Determine the thermal and humidity control targets for multiple chambers based on thermal and humidity compliance matrix;

[0030] Generate a candidate set of control strategies, including damper opening, cooling power and fan frequency, and the candidate set of control strategies satisfies thermal and humidity compliance matrix constraints.

[0031] Multi-chamber conflict identification is performed on the candidate set of control strategies. The conflict identification is as follows: each control strategy in the candidate set is substituted into the multi-chamber thermal-humidity network model to calculate the wet-bulb black-ball temperature index after implementing the strategy in each chamber. The wet-bulb black-ball temperature index after implementing the strategy in each chamber is compared with the target set of wet-bulb black-ball temperature indices for each chamber. The number of chambers exceeding the target threshold is counted. If the number of exceeding the threshold is greater than 0, the strategy is marked as a conflict strategy. Otherwise, it is marked as a non-conflict strategy, thus obtaining a subset of conflict strategies and a subset of non-conflict strategies.

[0032] The conflict strategy subset is optimized by a multi-chamber conflict resolution algorithm to obtain the resolved strategy subset;

[0033] The non-conflict strategy subset is merged with the resolved strategy subset to obtain the compliance control strategy set;

[0034] Based on the set of compliance control strategies and the set of partial order relations, a partial order lattice of control strategies is constructed; the partial order relation is defined as the Pareto priority of strategy A over strategy B, and the set of partial order relations is obtained by comparing all strategies in the set of compliance control strategies pairwise.

[0035] Furthermore, the construction of the partial-order lattice of the regulation strategy includes:

[0036] Each control strategy in the set of compliant control strategies is used as an element of the grid, and the element attributes include damper opening, cooling power, and fan frequency;

[0037] Define lattice operations, which include intersection and union operations. The intersection operation is to take the minimum value of the element attributes in two control strategies, and the union operation is to take the maximum value of the element attributes in two control strategies.

[0038] The cell base is determined as the minimum result of the intersection operation among all strategies, and the cell top is determined as the maximum result of the union operation among all strategies.

[0039] Furthermore, the grid path-guided multi-executor includes:

[0040] The target lattice path is selected from the partial order lattice of the control strategy by the lattice path priority evaluation algorithm. The target lattice path must simultaneously satisfy the safety margin and the comfort approximation efficiency constraints.

[0041] Initial control commands are generated based on the target grid path, and the parameters in the grid bottom strategy are converted into control signals for damper actuators, chiller unit actuators, and fan frequency converter actuators.

[0042] After sending the initial control command, collect temperature and humidity monitoring data of key areas and calculate the real-time wet-bulb black-bulb temperature index;

[0043] Based on the real-time wet-bulb black bulb temperature index, the control instructions are updated through closed-loop feedback adjustment rules, including maintaining the comfort range, optimizing the compliance range, and emergency retreat rules for exceeding limits.

[0044] Avoid action conflicts through collaborative timing control;

[0045] Built-in fault tolerance mechanism to reselect alternative grid paths in case of equipment failure or sensor malfunction.

[0046] Furthermore, the closed-loop feedback adjustment rules include:

[0047] Rule 1: If the real-time wet-bulb temperature index and relative humidity are within the comfort range, maintain the current control strategy and generate a maintenance command. The command content is the maintenance parameter of the current actuator control signal.

[0048] Rule 2: If the real-time wet-bulb black ball temperature index is in the compliant but not comfortable range, then move along the target grid path towards a more aggressive strategy;

[0049] Rule 3: If the real-time wet-bulb black sphere temperature index exceeds 34°C, a rollback command is immediately generated to roll back to the previous safe cell layer strategy on the target cell path, and the reason for exceeding the limit is recorded.

[0050] A system for constructing and dynamically controlling a windflow-thermal-humidity coupling model, used to implement the aforementioned method for constructing and dynamically controlling a windflow-thermal-humidity coupling model, the system comprising:

[0051] Data fusion and mesh generation module: used to collect multi-source temperature and humidity sensor data and wind network topology, fuse the multi-source temperature and humidity sensor data and wind network topology, generate and output thermal and humidity disturbance phase space mesh;

[0052] Compliance matte generation module: Based on the thermal and moisture perturbation phase space grid, the compliance evolution path is deduced through the thermal and moisture conduction matte generator, the compliance of the evolution path is verified, and the thermal and moisture compliance matte is output.

[0053] The partial order lattice generation module for control strategies is used to determine the control objectives of multi-chamber heat and humidity under the constraints of heat and humidity compliance matroid, generate and optimize the candidate set of control strategies, construct the partial order relationship between control strategies, and generate the partial order lattice of control strategies through the partial order lattice optimizer for multi-chamber conflict resolution.

[0054] Compared to existing technologies, the advantages of this invention are as follows: This invention accurately solves the problem that traditional methods struggle to characterize the thermal and humid environment in a multi-dimensional manner and are prone to conflict or exceeding limits under the strong coupling of multi-source disturbances (surrounding rock heat exchange, equipment heat dissipation, and surface climate cycle fluctuations) in deep mines. By fusing multi-source data and wind network topology through a spatiotemporal weighting dynamic weighting method, a three-dimensional thermal and humid disturbance phase space grid is constructed, preserving the triple spatiotemporal coupling effect and overcoming the limitations of traditional models in describing the environment in a one-sided way. A thermal and humid conduction matroid generator enumerates compliant evolution paths and constructs a matroid, overcoming the blind control caused by traditional single-path methods. A partial-order lattice optimizer for multi-chamber conflict resolution accurately resolves the problem of "one chamber compliant, multiple chambers exceeding limits," and also constructs a policy partial-order lattice containing Pareto dominance relations, addressing the pain point of unstructured traditional policies. This invention employs a lattice-path-guided multi-actuator system to achieve "safe and progressive" control and structured backoff, avoiding parameter oscillations and ensuring that temperature and humidity do not exceed 34°C and approach the comfort range of 26°C / 85%, thus guaranteeing the safety of downhole operations and the stability of electromechanical equipment. The invention utilizes matroid theory to systematically enumerate compliant evolution paths for complex coupled industrial systems, constructs a partially ordered lattice to achieve structured optimization of control strategies under multiple constraint conflicts, and designs a lattice-path-guided multi-actuator progressive closed-loop mechanism. This provides an innovative engineering paradigm that combines completeness, backoffability, and robustness for addressing industrial control scenarios with strong coupling of multiple variables and conflicts of multiple objectives. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 This is a flowchart of a method for constructing and dynamically controlling an airflow-thermal-humidity coupling model in this invention;

[0057] Figure 2 This is a hierarchical grid structure and Pareto dominance diagram of the partial order lattice for the control strategy in this invention;

[0058] Figure 3 This is a functional block diagram of a system for constructing and dynamically controlling a coupled airflow, heat, and humidity model in this invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] Example 1:

[0061] Please see Figure 1 As shown, this embodiment provides a method for constructing and dynamically controlling a wind-flow-thermal-humidity coupling model, including:

[0062] Step S10: Collect multi-source temperature and humidity sensor data and wind network topology, fuse the multi-source temperature and humidity sensor data and wind network topology, and generate and output the thermal and humidity disturbance phase space grid.

[0063] The multi-source temperature and humidity sensing data includes four core data categories: temperature data, humidity data, wind speed data, and heat source intensity data. The temperature data refers to the real-time temperature values ​​of spatial nodes in each monitoring area within the mine. This data is collected by distributed temperature and humidity sensors deployed in key areas such as the mining face, central pump room, and intake shaft. The collection frequency is determined through variance analysis of historical heat and humidity fluctuation data; a larger variance results in a higher collection frequency. The humidity data refers to the real-time relative humidity values ​​of each spatial node, collected synchronously with the temperature data through distributed temperature and humidity sensors. The value range is the historical measured relative humidity range within the mine, determined by statistically analyzing the maximum and minimum values ​​of mine humidity data over the past three years. The wind speed data refers to the real-time airflow velocity within each ventilation path. The temperature is collected by a wind speed sensor installed at the center of the airflow cross-section. During the collection, the vortex region within the airflow path must be avoided. The location of the vortex region is determined by CFD simulation of the mine airflow field. The heat source intensity data refers to the real-time heat generation power of the equipment during operation in the mine. It is calculated by combining the real-time current and voltage data collected by the equipment controller with the rated power of the equipment. The specific conversion formula is: Heat source intensity = Rated power of equipment × (Real-time current / Rated current) × (Real-time voltage / Rated voltage). The rated power of the equipment is obtained from the equipment technical parameter manual, and the real-time current and voltage are collected from the current and voltage sensors built into the equipment.

[0064] The ventilation network topology refers to the connection relationships and physical parameter set of each spatial node (including chambers, ventilation shafts, etc.) and airway in the mine ventilation network, which is constructed by combining the mine ventilation system design drawings with on-site measured data. The physical parameters included in the ventilation network topology are airway length, airway cross-sectional area, and airway resistance coefficient. The airway length is obtained through on-site measurement using a laser rangefinder. The airway cross-sectional area is calculated by measuring the length and width of the rectangular cross-section or the diameter of the circular cross-section of the airway. The airway resistance coefficient is calculated by back-calculating the ventilation resistance formula h=RQ² using the wind speed and wind pressure data collected within the airway, where h is the wind pressure collected by a wind pressure sensor; Q is the air volume calculated by multiplying the wind speed and the airway cross-sectional area; and R is the airway resistance coefficient.

[0065] Specifically, the fusion processing of multi-source temperature and humidity sensor data with the wind network topology is as follows:

[0066] The first step involves spatiotemporal alignment of multi-source temperature and humidity sensor data. First, a unified time step is determined. The length of the time step is determined by referencing the seasonal periodic fluctuations of the mine's surface climate, with the core principle being the ability to comprehensively capture the transmission effects of different seasons (such as high temperature and humidity in summer and low temperature and humidity in winter) on the underground thermal and humidity environment. This is determined by combining the periodic variation cycles of temperature and humidity in the mine's historical thermal and humidity data from the past five years, ensuring that the time step covers both the climatic cycle characteristics and meets the data timeliness requirements for real-time control. Next, the criteria for dividing spatial nodes within the mine are clarified. Taking key operation and equipment areas such as the mining face, central pump room, intake shaft, and return airway as the core, and combining the airflow transmission paths in the ventilation network topology, the range of spatial nodes to be monitored is determined. Each spatial node corresponds to a specific physical location within the mine. For multi-source temperature and humidity sensor data, according to a predetermined unified time step, the data collected by different sensors at different times are matched to the corresponding spatial nodes: If there are similar data collected by multiple sensors in the same spatial node within the same time step (such as temperature data collected by sensors at different locations in the same area), the data is filtered and fused based on the physical distance between the sensor and the center of the spatial node and the accuracy of the sensor's historical data, and finally the temperature, humidity, wind speed and heat source intensity data of each spatial node at each time step are obtained, i.e., spatiotemporally aligned data.

[0067] The second step is to calculate the spatiotemporal weights. The time weights are determined based on the time difference between the data acquisition time and the current time, as well as the seasonal cyclical variation. Specifically, an exponential decay model combined with a seasonal cosine correction factor is used; the larger the time difference, the more significant the weight decay. Simultaneously, based on the differences in heat and moisture conduction characteristics in different seasons such as summer and winter, a seasonal correction coefficient is obtained through regression analysis of historical data to ensure that the weights reflect the impact of climate cycles on data timeliness. The spatial weights are determined based on the straight-line distance between the sensor and the spatial node, and the wind resistance coefficient in the wind network topology. Based on an inverse distance weighting method and incorporating a wind resistance correction term, the closer the sensor is to the spatial node, the greater the weight. Simultaneously, the wind resistance coefficient in the wind network topology for that region is referenced; the greater the resistance, the stronger the obstruction of heat and moisture transfer by the airflow. The corresponding spatial weights need to be adjusted inversely proportionally to the proportion of the resistance coefficient.

[0068] The third step is to perform weighted processing on the spatiotemporally aligned data. The temperature, humidity, wind speed, and heat source intensity in the spatiotemporally aligned data are multiplied by their corresponding time weights w_t and spatial weights w_s, respectively, to obtain weighted temperature data, weighted humidity data, weighted wind speed data, and weighted heat source intensity data.

[0069] The fourth step is to construct a thermal and humidity perturbation phase space grid. Weighted temperature data, weighted humidity data, weighted wind speed data, and weighted heat source intensity data are used as quadruples of perturbation dimensions to finally form a three-dimensional tensor structure thermal and humidity perturbation phase space grid, with the tensor dimension being spatial nodes × time steps × perturbation dimension.

[0070] In the three-dimensional tensor structure of the thermal and humid disturbance phase space grid, the spatial node dimension covers all key areas within the mine that need to be monitored, such as the mining face, central pump room, intake shaft, and return air shaft. The specific number of nodes is determined based on the actual scale of the mine and monitoring requirements. The time step length is divided with reference to the seasonal periodic fluctuations of the mine's surface climate to ensure that the impact of climate cycle changes on the mine's thermal and humid environment can be fully captured. The four elements of the disturbance dimension correspond to temperature, humidity, wind speed, and heat source intensity, respectively. The value range of each element is determined based on historical measured data of the mine and equipment technical parameters to ensure the engineering practicality of the data.

[0071] This step constructs a thermal and humidity disturbance phase space grid by collecting multi-source temperature and humidity sensor data and wind network topology. Its core advantage lies in its ability to simultaneously preserve the triple spatiotemporal coupling effect of surrounding rock thermal inertia, instantaneous thermal shock from equipment, and climate cycle conduction. Traditional grid models can only depict spatial distribution or temporal variation in a single way and cannot simultaneously bear the connectivity of the wind network topology. However, this step fuses multi-source temperature and humidity sensor data and wind network topology to balance the spatiotemporal distribution differences of instantaneous thermal shock from equipment, solving the technical problem that traditional models cannot effectively depict the multi-dimensional coupling of the mine's thermal and humidity environment. Furthermore, this thermal and humidity disturbance phase space grid supports the calculation of local heat conduction differential operators and the projection of global climate modes, providing a complete and accurate input data foundation for the subsequent step S20 to deduce the thermal and humidity compliance evolution path. This ensures the compliance and robustness of the subsequent evolution path and provides precise initial data support for dynamic control of mine thermal and humidity.

[0072] Step S20: Based on the thermal and moisture perturbation phase space grid, deduce the compliance evolution path through the thermal and moisture conduction matte generator and output the thermal and moisture compliance matte.

[0073] The thermal and moisture conduction matroid generator is an algorithm module used to extract the evolution law of thermal and moisture states from the thermal and moisture perturbation phase space grid and generate a set of paths that meet compliance conditions based on the law. Its core function is to transform the evolution process of thermal and moisture states into the construction process of independent sets of matroids by introducing the matroid theory framework, thereby realizing the systematic enumeration of all possible compliant paths.

[0074] Specifically, the thermal and moisture conduction matte generator includes the following steps:

[0075] Based on the thermal and humidity perturbation phase space grid, the thermal and humidity state parameters of each spatial node at different time steps need to be extracted first. These parameters include a quadruple of temperature, humidity, wind speed, and heat source intensity, which are directly obtained from the perturbation dimension of the thermal and humidity perturbation phase space grid. By performing spatiotemporal correlation analysis on these parameters, a transition rule for the thermal and humidity state is established. This transition rule describes how the thermal and humidity state at a certain time step evolves into the thermal and humidity state at the next time step. Specifically, the rules for the transfer of thermal and humidity states are established as follows: For any spatial node, its temperature value at time step t+1 is calculated from the temperature value, humidity value, wind speed value, and heat source intensity value at time step t using the heat conduction equation. The heat conduction equation is: T(t+1) = T(t) + k1 × (Theat source - T(t)) × Δt + k2 × (Tneighborhood - T(t)) × Δt × vwind, where k1 is the heat source conduction coefficient, obtained by fitting experimental data of equipment heat dissipation characteristics; k2 is the spatial conduction coefficient, calculated based on the wind path resistance coefficient in the wind network topology; Δt is the time step interval; Tneighborhood is the temperature value of adjacent spatial nodes; vwind is the wind speed value; T(t+1) is the temperature value at time t+1; and Theat source is the heat source intensity value. The transfer rules for humidity value, wind speed value, and heat source intensity value are established using a similar method, based on the humidity diffusion equation, the fluid dynamics conservation equation, and the equipment power change curve, respectively.

[0076] Using the aforementioned transition rules, starting with the initial time-step thermal and humidity state in the thermal and humidity perturbation phase space grid, all possible future thermal and humidity state evolution paths are generated. Each evolution path consists of a series of continuous thermal and humidity perturbation phase space grid slices, with each slice corresponding to the thermal and humidity state of one time step. The generation process employs a breadth-first search strategy, expanding the thermal and humidity state of each time step until a preset maximum prediction duration is reached. The maximum prediction duration is determined based on the decision-making cycle of mine thermal and humidity control, typically ranging from 24 to 72 hours.

[0077] Each generated evolution path undergoes compliance verification to obtain compliant evolution paths. The core of compliance verification is to determine whether the temperature of all spatial nodes in the path meets the red line requirements of 30℃ and 34℃ at all time steps. Specifically, for critical areas such as mining faces and central pump rooms, the temperature must not exceed 30℃; for non-operational areas such as intake shafts and return air tunnels, the temperature must not exceed 34℃. Simultaneously, the wet-bulb black-bulb temperature index (WBI) needs to be calculated based on humidity and wind speed. The WBI is calculated using the formula in the national standard: WBI = 0.7 × wet-bulb temperature + 0.2 × black-bulb temperature + 0.1 × dry-bulb temperature. The wet-bulb temperature is obtained by converting temperature and humidity using thermodynamic formulas, while the black-bulb temperature is obtained by correcting for temperature and wind speed. When the WBI exceeds the corresponding safety threshold, the path is deemed non-compliant.

[0078] Based on the compliance evolution path, a thermal and humidity compliance matte is constructed using a matte construction algorithm. Specifically, the matte construction algorithm is as follows:

[0079] The first step is to define the base set of the thermal and humidity compliance matroid. The base set is the collection of all compliance evolution paths that have passed compliance verification; that is, each element in the base set is a compliance evolution path.

[0080] The second step is to define the family of independent sets. A family of independent sets consists of subsets of the fundamental set that satisfy the matroid axiom. Specifically, the heritability axiom states that if a set of compliant evolutionary paths is an independent set, then all its subsets are also independent sets. The commutativity axiom states that if two independent sets A1 and B1 exist, and A1 has fewer elements than B1, then there exists a compliant evolutionary path that belongs to B1 but not to A1; adding this path to A1 results in a new set that is still an independent set.

[0081] The third step is to determine the matroid basis. The matroid basis is the independent set with the largest number of elements in the family of independent sets. That is, the matroid basis is the set consisting of the longest compliant evolution path. The longest compliant evolution path is the evolution path that remains compliant throughout the preset maximum prediction duration.

[0082] The fourth step is to verify the matroid properties. By traversing all elements in the family of independent sets, we verify whether they satisfy the axioms of heritability and commutativity, ensuring that the constructed thermo-humidity compliant matroid conforms to the mathematical definition of a matroid.

[0083] The structure of a thermo-humidity compliant matroid consists of a foundation set, a family of independent sets, and a matroid basis. The size of the foundation set depends on the number of compliant evolutionary paths; a larger number indicates more possibilities for compliant evolution under the current thermo-humidity perturbation conditions. The family of independent sets uses the axioms of heredity and commutativity to structurally organize the compliant evolutionary paths, allowing operations on the path set to follow the mathematical rules of matroid theory. The matroid basis, as the set of the longest compliant evolutionary paths, provides a benchmark reference for subsequent optimization of control strategies.

[0084] This step transforms the thermal and moisture perturbation phase space grid into a thermal and moisture compliance matte through a thermal and moisture conduction matte generator. Its core function is to enumerate all possible compliance evolution paths, providing a comprehensive decision-making basis for dynamic thermal and moisture control in mines. Compared with traditional thermal and moisture prediction models, this step has three significant advantages: First, the matte structure enables a structured description of the set of compliance paths, overcoming the limitation of traditional models that can only output a single path. This allows the selection of control strategies to be based on multi-path comparisons, improving the robustness of decision-making. Second, the heritability of the matte ensures that all sub-paths of any compliance path are compliant. This means that during the control process, any segment of the path can be extracted as a reference according to actual needs, enhancing the flexibility of path application. Third, the commutativity of the matte ensures that local strategies between different compliance paths can be interchanged without compromising overall compliance. This provides a mathematical basis for the optimized combination of control strategies in subsequent steps, enabling local control measures to be effectively adapted globally. The thermal and humid compliance matte generated in this step can fully preserve all compliance evolution possibilities under strong perturbation conditions, providing key support for solving the control problem caused by strong nonlinearity and multivariate coupling in the thermal and humid environment of mines. This enables the subsequent generation of control strategies to be optimized on the basis of fully covering the compliance space, avoiding control misconduct caused by incomplete path enumeration.

[0085] For example, suppose that in a thermal and humidity disturbance phase space grid of a certain mine, the initial time step has a mining face temperature of 28℃, humidity of 80%, wind speed of 1.5m / s, and heat source intensity of 50kW. The thermal and humidity evolution paths for the next 24 hours are generated using the state transition rules of the thermal and humidity conduction matroid generator. During compliance verification, all paths with mining face temperatures exceeding 30℃ are removed, retaining 10 compliant evolution paths. Using a matroid construction algorithm, these 10 paths constitute the basis set of the thermal and humidity compliance matroid, with the three paths of length 24 hours forming the matroid basis. Based on this thermal and humidity compliance matroid, subsequent steps can extract the optimal evolution path as the control target, ensuring that the control strategy keeps the mine's thermal and humidity environment within the compliance range.

[0086] Step S30: Under the thermal and humidity compliance matrix constraint, generate a partial order lattice for the control strategy through a partial order lattice optimizer for multi-chamber conflict resolution.

[0087] The partial-order lattice optimizer for multi-chamber conflict resolution refers to an algorithm module that uses the compliance evolution path in the thermal and humidity compliance matte as the constraint boundary. It addresses the differences in thermal and humidity control requirements among different chambers such as the central pump room, mining face, and intake shaft. First, it identifies conflicts between control strategies, then optimizes the strategy set through a conflict resolution algorithm, and finally constructs a control strategy lattice with an algebraic structure based on partial-order relations. Its core function is to resolve the conflict problem of multi-chamber control objectives and provide an evolvable and regressible structured organization method for control strategies.

[0088] Specifically, the partial order lattice optimizer for multi-chamber conflict resolution includes:

[0089] First, thermal and humidity control targets for multiple chambers are determined based on a thermal and humidity compliance matte model. Each independent set in the thermal and humidity compliance matte model corresponds to a compliance evolution path, and each compliance evolution path contains the target values ​​of temperature, humidity, and wind speed for each chamber at different time steps. These target values ​​are then converted into target thresholds for the wet-bulb black-bulb temperature index for each chamber. The wet-bulb black-bulb temperature index is calculated using the industry standard formula: Wet-bulb black-bulb temperature index = 0.7 × wet-bulb temperature + 0.2 × black-bulb temperature + 0.1 × dry-bulb temperature. The wet-bulb temperature is obtained by converting the temperature and humidity in the path using the thermodynamic balance equation, and the black-bulb temperature is obtained by correcting the temperature and wind speed in the path using the convective heat transfer coefficient. The convective heat transfer coefficient is calculated based on the cross-sectional area of ​​the wind path and the wind speed in the wind network topology. Through this process, a target set of wet-bulb black-bulb temperature indices for multiple chambers is obtained, which provides a benchmark for the generation of subsequent control strategies.

[0090] Secondly, a candidate set of control strategies is generated. The control strategies are based on a triplet consisting of damper opening, cooling power, and fan frequency. Each unit must satisfy the constraints of a thermal and humidity compliance matrix, meaning that the thermal and humidity state of each chamber must fall within the compliance evolution path after the strategy is implemented. Specifically, the value range and calculation method of each parameter are as follows:

[0091] (1) Damper opening: The value ranges from 0% (fully closed) to 100% (fully open). Its calculation is based on the wind resistance coefficient in the wind network topology and the wind speed target value in the compliance evolution path, using the ventilation volume balance equation Q=K×α×S× In reverse calculation, Q is the target air volume, obtained by multiplying the target wind speed value by the cross-sectional area of ​​the air path; K is the flow coefficient, obtained by fitting historical ventilation data of the air network; α is the damper opening; S is the cross-sectional area of ​​the air path; ΔP is the pressure difference between the two ends of the air path, calculated from the fan pressure data in the air network topology; R is the air path resistance coefficient.

[0092] (2) Cooling power: The value range is from the rated minimum power to the rated maximum power of the equipment. It is calculated based on the difference between the temperature target value and the current temperature in the compliance evolution path and the heat source intensity. The heat balance equation is P=c×m×ΔT / Δt2+Pheat source, where c is the specific heat capacity of air, which is taken as the industry standard value; m is the air mass of the chamber, which is calculated from the chamber volume and air density; ΔT is the temperature difference; Δt2 is the control time step; and Pheat source is the heat source intensity in the path.

[0093] (3) Fan frequency: The value range is from the rated minimum frequency to the rated maximum frequency of the equipment. Its calculation is based on the wind speed target value and the fan characteristic curve in the compliance evolution path. The fan characteristic curve is obtained by fitting the performance parameters provided by the fan manufacturer with the wind speed-frequency data measured on site.

[0094] Based on the above parameter calculation method, the range of values ​​of each parameter is traversed to generate a candidate set of control strategies that satisfy the thermal and humidity compliance matrix constraints. This set contains all triplet strategies that may make the thermal and humidity state of the multi-chamber fall within the compliance evolution path.

[0095] Subsequently, multi-chamber conflict identification is performed on the candidate set of control strategies. Conflict is defined as follows: if the implementation of a certain control strategy can bring the wet-bulb black-ball temperature index of some chambers to the target threshold, but causes the wet-bulb black-ball temperature index of at least one other chamber to exceed the target threshold, then the strategy conflicts with the multi-chamber control objective. Specifically, the conflict identification method is as follows: each strategy in the candidate set of control strategies is substituted into the multi-chamber thermal-humidity network model to calculate the wet-bulb black-ball temperature index of each chamber after implementing the strategy; the wet-bulb black-ball temperature index of each chamber after implementing the strategy is compared with the target set of wet-bulb black-ball temperature indices of each chamber, and the number of chambers exceeding the target threshold is counted; if the number of exceeding the threshold is greater than 0, the strategy is marked as a conflicting strategy; otherwise, it is marked as a non-conflicting strategy. Through this process, a subset of conflicting strategies and a subset of non-conflicting strategies are obtained. The multi-chamber thermal and humidity network model uses the control volume method to discretize the chambers and calculates the heat and humidity transfer through the air volume distribution of the ventilation network, which is a common modeling method for thermal and humidity control in mines.

[0096] Next, the conflict strategy subset is optimized using a multi-chamber conflict resolution algorithm to obtain the resolved strategy subset. Specifically, the multi-chamber conflict resolution algorithm is as follows:

[0097] The first step is to determine the weight coefficients of each chamber. Based on the operational importance and thermal and humidity sensitivity of each chamber, the weight coefficient w_i of each chamber is calculated using the analytic hierarchy process (AHP), where i is the chamber number. The sum of the weight coefficients of all chambers is 1. The rationality of the weight coefficients is verified through a consistency check.

[0098] The second step is to calculate the conflict degree C1 of the conflict strategy. The conflict degree is defined as the weighted sum of deviations caused by the strategy that result in the wet-bulb black-bulb temperature index of each chamber exceeding the target threshold. The calculation formula is C1=Σw_i×|T_i-T_i0| / T_i0, where T_i is the wet-bulb black-bulb temperature index of chamber i after the strategy is implemented, and T_i0 is the target threshold of chamber i. The higher the conflict degree, the more serious the conflict of the strategy.

[0099] The third step is to adjust the strategy parameters. Conflicting strategies are sorted from highest to lowest conflict level, with priority given to strategies with higher conflict levels. The adjustment rules are as follows: If a chamber's wet-bulb temperature index exceeds the standard due to strategy parameters, for the damper opening, if the chamber is on the intake side, increase the opening to increase airflow; if it is on the return side, decrease the opening to reduce the inflow of hot air. For the cooling power, if the chamber's temperature is too high, increase the cooling power; if the humidity is abnormal, adjust it in conjunction with the dehumidification module. The dehumidification module power is linked to the cooling power, and the linkage coefficient is determined through a temperature and humidity coupling experiment. For the fan frequency, if the chamber's airflow is insufficient, increase the frequency to increase airflow; otherwise, decrease the frequency.

[0100] The fourth step is to verify the adjusted strategy. The adjusted strategy is then substituted back into the multi-chamber thermal-humidity network model to calculate the wet-bulb black-bulb temperature index for each chamber. If all chambers meet the target threshold, the strategy is included in the subset of strategies after resolution. If there are still cases of exceeding the target, steps three and four are repeated until the strategy is compliant or the maximum number of adjustments is reached. The maximum number of adjustments is determined by historical adjustment efficiency data.

[0101] The non-conflict strategy subset is merged with the resolved strategy subset to obtain the compliant control strategy set. Based on this set, a partial order relation is constructed among the control strategies. The partial order relation is defined as Pareto dominance: if strategy A and strategy B both belong to the compliant control strategy set, and for all chambers, the wet-bulb black-ball temperature index T_Ai after implementing strategy A is not inferior to that of strategy B, i.e., T_Ai < T_Bi (cooling) or T_Ai ≥ T_Bi (heating), the specific direction is determined by the target set of wet-bulb black-ball temperature indices for multiple chambers, and at least one chamber has a wet-bulb black-ball temperature index superior to that of strategy B, then strategy A is said to be Pareto superior to strategy B, denoted as A ≥ B. The partial order relation set is obtained by comparing all strategies in the compliant control strategy set pairwise.

[0102] Finally, based on the set of compliance control strategies and the set of partial order relations, a partial order lattice of control strategies is constructed. Specifically, the method for constructing the partial order lattice structure of control strategies is as follows:

[0103] The first step is to determine the elements of the grid. Each strategy in the set of compliant control strategies is used as an element of the grid, and the attributes of the element are the specific values ​​of damper opening, cooling power, and fan frequency.

[0104] The second step is to define the grid operations. Intersection operation (conservative superposition): For any two strategies A and B, the intersection operation result is strategy C. The parameters of strategy C are taken as the minimum values ​​among the corresponding parameters of A and B, such as the damper opening being min(α_A, α_B), the cooling power being min(P_A, P_B), and the fan frequency being min(f_A, f_B). This ensures that the intersection operation result is a more conservative strategy, guaranteeing control safety. Here, α_A and α_B are the damper opening parameters in strategies A and B, respectively; P_A and P_B are the cooling power parameters in strategies A and B, respectively; and f_A and f_B are the fan frequency parameters in strategies A and B, respectively. Union operation (aggressive coordination): For any two strategies A and B, the union operation result is strategy D. The parameters of strategy D are taken as the maximum values ​​among the corresponding parameters of A and B. This ensures that the union operation result is a more aggressive strategy, pursuing a better thermal and humidity comfort state.

[0105] The third step is to determine the vertices and paths of the lattice. The lattice base is the minimum result of the intersection operation among all policies, i.e., the most conservative policy, where each parameter value is minimized; the lattice vertex is the maximum result of the union operation among all policies, i.e., the most aggressive policy, where each parameter value is maximized; the set of lattice vertices forms the Pareto optimal front, meaning no other policy can Pareto outperform these vertex policies; the lattice path is a sequence of policies that progressively increase the regulatory strength from the lattice base to the lattice vertex along a partial order relation, with each path corresponding to a regulatory strength gradient. An exemplary structure is shown below. Figure 2 In the diagram, the arrows indicate Pareto dominance, and the small black rectangles mark the Pareto optimal front nodes.

[0106] The structural characteristics of the partial-order lattice of control strategies are as follows: elements form an ordered hierarchy through Pareto dominance relations; lattice operations support safe merging and aggressive optimization of strategies; and lattice paths provide traceable and reversible control directions. The number of elements depends on the size of the set of compliant control strategies; a larger size indicates more control schemes to choose from under the thermal and humidity compliant matroid constraint. The density of the partial-order relations reflects the differences in quality between strategies; a higher density indicates a higher precision in strategy optimization.

[0107] This step generates a partial-order lattice of control strategies using a partial-order lattice optimizer for multi-chamber conflict resolution. Its core function is to resolve the objective conflict problem in multi-chamber thermal and humidity control and provide a structured strategy system for subsequent dynamic control. Compared to traditional multi-objective optimization methods, it has three core advantages: First, it overcomes the limitation of traditional methods outputting unstructured point sets. Through the algebraic structure of the partial-order lattice, it achieves ordered organization of strategies, making the superiority / inferiority relationships, merging rules, and evolution paths between strategies clearly discernible, thus avoiding blind control decisions. Second, the multi-chamber conflict resolution algorithm combines chamber weights with… The thermal and humidity compliance matrix constraint ensures that the resolved strategy meets the control requirements of the core chamber while remaining within the compliance evolution path, solving the "one-sided" problem in traditional conflict resolution. Thirdly, lattice operations and lattice path design are naturally suited to the engineering logic of "safety first, gradual trial and error" in downhole control. The conservative superposition of intersection operations avoids the risk of exceeding limits caused by aggressive control, and the gradient evolution of the lattice path supports gradual control from conservative to aggressive. Furthermore, when exceeding limits, it can retreat along the path to a safe strategy, a structured safety retreat mechanism lacking in traditional PID or MPC control. The partial-order lattice of the control strategy generated in this step provides a clear strategy selection space and evolution direction for the dynamic closed-loop control in step S40, ensuring that the temperature and humidity in key areas remain stable within the compliant comfort range while balancing the safety and economy of control.

[0108] For example, suppose a mine contains three key chambers: a mining face, a central pump room, and an intake shaft. The target thresholds for the wet-bulb black-sphere temperature index, determined by thermal-humidity compliance simulation, are 28℃, 30℃, and 32℃, respectively. The generated control strategy candidate set contains 5 strategies. Among them, strategy 1 (damper opening 60%, cooling power 80kW, fan frequency 45Hz) was marked as a conflict strategy after its implementation, resulting in a mining face temperature of 27℃ (compliant), a central pump room temperature of 31℃ (exceeding the standard), and an intake shaft temperature of 31℃ (compliant). Through the multi-chamber conflict resolution algorithm, with a weight of 0.4 for the central pump room, 0.5 for the mining face, and 0.1 for the intake shaft, the conflict degree of strategy 1 was 0.4×|31-30| / 30+0.5×|27-28| / 28≈0.023. Adjusting the cooling power of strategy 1 to 85kW, we obtained the resolved strategy 1'. After its implementation, the temperatures of the three chambers were 26.5℃, 29.8℃, and 31.5℃ (all compliant). Non-conflict strategies 2-5 are merged with the resolved strategy 1' into a set of compliant control strategies. Based on Pareto dominance judgment, strategy 3 (70% damper opening, 90kW cooling power, 50Hz fan frequency) is superior to strategy 1' (lower temperatures in all three chambers). Strategy 5 (50% damper opening, 70kW cooling power, 40Hz fan frequency) is the bottom of the grid (most conservative), and strategy 3 is the top of the grid (most aggressive). Strategy 3 and strategy 4 (65% damper opening, 85kW cooling power, 48Hz fan frequency) constitute the Pareto optimal frontier, ultimately forming a partial-order grid of control strategies.

[0109] Step S40: Based on the partial order grid of the control strategy, execute the dynamic closed-loop control command through the multi-actuator guided by the grid path.

[0110] The grid-path guided multi-actuator refers to a control module that coordinates the actions of three types of actuators—damper actuators, refrigeration unit actuators, and fan frequency converter actuators—using the grid path in the partial-order grid of the control strategy as the control direction, and combines real-time monitoring of key area thermal and humidity data to achieve closed-loop feedback adjustment. Its core function is to execute the control strategy according to the principle of "safe and gradual" implementation, while also possessing an over-limit retreat mechanism to ensure that the absolute red line of 34℃ is not exceeded and that the target of 26℃ / 85% comfort level is gradually approached.

[0111] Specifically, the grid path-guided multi-executor steps include:

[0112] First, target lattice paths are selected from the partially ordered lattice of the control strategy. The target lattice paths must simultaneously meet the requirements of safety margin and comfortable approximation efficiency, specifically determined by a lattice path priority evaluation algorithm. Specifically, the lattice path priority evaluation algorithm is as follows:

[0113] The first step is to calculate the safety margin for each grid path. The safety margin is defined as the minimum difference between the wet-bulb black-bulb temperature index of the critical region corresponding to all strategies on the path and the 34℃ red line. The calculation formula is M_s = min(34 - T_wbgt(s)), where s is the control strategy on the path, T_wbgt(s) is the wet-bulb black-bulb temperature index after strategy s is implemented, and M_s is the safety margin. A larger safety margin indicates a stronger resistance to disturbances on the path. This parameter is calculated by extracting historical implementation data of each strategy in the partial-order grid of control strategies. When historical data is insufficient, simulation using a heat and humidity conduction model is employed.

[0114] The second step is to calculate the comfort approximation efficiency for each grid path. Comfort approximation efficiency is defined as the rate at which the wet-bulb black-bulb temperature index approaches 26°C from its initial value as the path moves from the bottom to the top of the grid. The formula is E_c = (T_init - 26) / N, where T_init is the wet-bulb black-bulb temperature index corresponding to the bottom strategy, N is the number of strategy steps required to reach 26°C for the first time from the bottom of the grid, and E_c is the comfort approximation efficiency. Higher efficiency indicates a faster attainment of the comfort goal. N is determined by traversing the strategy sequence along the grid path.

[0115] The third step is to determine the target grid path. A weighted summation method is used to calculate the priority score S1 for each path. The score formula is S1 = w1 × M_s + w2 × E_c, where w1 is the safety margin weight and w2 is the comfort approximation efficiency weight. w1 and w2 are determined using the analytic hierarchy process (AHP), with priority given to ensuring that w1 > w2. The path with the highest score is selected as the target grid path.

[0116] Based on the target grid path, an initial control command is generated. The initial control command corresponds to the grid-based strategy of the target grid path, which is the most conservative strategy. The damper opening, cooling power, and fan frequency parameters in the grid-based strategy need to be converted into control signals for three types of actuators:

[0117] (1) Control signal of damper actuator: The damper actuator is an electric proportional actuator. The control signal is PWM duty cycle. The correspondence between duty cycle and damper opening is determined by actuator calibration experiment. The expression is D=a×α+b, where D is the duty cycle, α is the damper opening, and a and b are calibration coefficients. The result is obtained by fitting duty cycle data under different openings.

[0118] (2) Control signal of refrigeration unit actuator: Refrigeration power is achieved by adjusting the compressor speed. The control signal is the speed command. The relationship between speed and refrigeration power is determined based on the unit characteristic curve. The characteristic curve is obtained by correcting the power-speed data provided by the manufacturer and the actual measured data on site. The expression is n=e×P_cool+f, where n is the compressor speed, P_cool is the refrigeration power, and e and f are characteristic coefficients.

[0119] (3) Control signal of the wind turbine frequency converter: The wind turbine frequency is adjusted by the frequency converter. The control signal is a frequency command, which directly corresponds to the wind turbine frequency parameter in the strategy. The frequency command range is consistent with the rated frequency of the frequency converter, and the value is the calculated value of the wind turbine frequency in the strategy.

[0120] Initial control commands are sent to the corresponding actuators. After the actuators act according to the commands, they collect temperature and humidity monitoring data for key areas. Key areas include the mining face, central pump room, and air intake shaft. Temperature and humidity monitoring data are collected by distributed temperature and humidity sensors deployed in each area. The collection frequency is consistent with the control time step. The control time step is determined by the strategy steps of the target grid path and the control cycle. The control cycle is experimentally determined based on the dynamic response speed of the thermal and humidity environment. Simultaneously, the real-time wet-bulb / black-bulb temperature index T_wbgt is calculated based on the collected temperature and humidity monitoring data. The calculation method is the same as in step S30, i.e., T_wbgt = 0.7 × T_wet + 0.2 × T_black + 0.1 × T_dry, where T_wet is the wet-bulb temperature, T_black is the black-bulb temperature, and T_dry is the dry-bulb temperature, all of which are converted or corrected from the real-time temperature and humidity data.

[0121] Based on the real-time wet-bulb spherical temperature index, the control command is updated through a closed-loop feedback adjustment rule. Specifically, the closed-loop feedback adjustment rule is as follows:

[0122] Rule 1: Comfort Range Maintenance. If the real-time wet-bulb sphere temperature index is 26℃±1℃ and the relative humidity is 85%±3% (comfort range), then maintain the current control strategy and generate a maintenance command. The command content is the maintenance parameters of the current actuator control signal.

[0123] Rule 2, Compliance Range Optimization. If the real-time wet-bulb black bulb temperature index is between 27℃ and 34℃ (compliant but not in the comfort range), then move towards a more aggressive strategy along the target grid path. The step size is determined by the temperature and humidity deviation, and the step size formula is Δs=k×(T_target-T_real), where Δs is the number of strategy steps, T_target is the 26℃ comfort target, T_real is the real-time wet-bulb black bulb temperature index, and k is the step size coefficient, which is determined by fitting the deviation-step size relationship of historical control data to ensure that the temperature and humidity changes are stable and without abrupt changes after each adjustment.

[0124] Rule 3, Emergency Retreat for Exceeding Limits. If the real-time wet-bulb black bulb temperature index exceeds 34°C, a retreat command is immediately generated, reverting to the previous safe grid layer strategy on the target grid path. At the same time, the reason for exceeding the limit, such as sudden thermal shock of the equipment or sensor malfunction, is recorded for subsequent optimization of the target grid path.

[0125] During the update of control commands, coordinated timing control of multiple actuators avoids action conflicts. For example, when adjusting the fan frequency, the frequency change signal must be sent to the fan frequency converter actuator 0.5 control time steps in advance. After the fan speed stabilizes (the stabilization time is determined through airflow response experiments in the air network), the damper opening and cooling power are adjusted to avoid temperature and humidity fluctuations caused by sudden airflow changes. The coordinated timing is determined by an actuator action timing table, which is formulated based on the principle of minimizing interference from multi-actuator linkage experiments, ensuring that the superposition effect of each actuator's actions meets the control expectations.

[0126] In addition, to cope with equipment failures or sensor anomalies, the actuator has a built-in fault tolerance mechanism. When an actuator reports a fault signal, such as a stuck damper or a refrigeration unit shutdown, the system immediately selects and excludes the faulty actuator parameters from the partial order grid of the control strategy, re-determines the new target grid path according to the grid path priority evaluation algorithm, and generates control instructions adapted to the spare path. When sensor data is abnormal (such as data fluctuations exceeding a threshold, which is determined by the sensor accuracy level and the standard deviation of historical data), a multi-sensor data fusion algorithm is used to obtain reliable temperature and humidity data, ensuring that the closed-loop feedback is not based on abnormal data. The fusion weight of the multi-sensor data fusion algorithm is determined based on the reliability coefficient of the sensor, which is calculated through sensor calibration records.

[0127] This step achieves dynamic closed-loop control through multiple actuators guided by lattice paths. Its core function is to transform the structured strategy in the partial-order lattice of the control strategy into physical execution actions, while ensuring the safety and effectiveness of the control through real-time feedback. Compared to traditional PID control or model predictive control (MPC), this step has three unique advantages: First, a structured safety backoff mechanism. Traditional control can only avoid exceeding limits by limiting parameters, while this step, based on the grid structure of the partial-order grid of the control strategy, can directly backoff to a verified safety strategy. The backoff logic is clearer and the anti-disturbance capability is stronger, completely solving the problem of "parameter oscillation after exceeding limits" in traditional control. Second, a progressive control logic. By controlling the step size of the target grid path, a smooth transition from conservative to aggressive control is achieved, avoiding parameter mutations caused by rapid approach to the target in traditional control. This is in line with the characteristics of large inertia and slow response in the hot and humid environment downhole, reducing the losses from frequent equipment start-ups and shutdowns. Third, multi-actuator collaborative timing optimization. Traditional control often uses parallel actions, which are prone to control deviations due to interference between actuators. This step, through timing coordination, enables the actions of each actuator to form a complementary effect, improving control accuracy. For example, the fan air volume is adjusted first to establish the airflow foundation, and then the cooling power is adjusted to control the temperature, avoiding the waste of cooling power when the air volume is insufficient. This step enables the stabilization of temperature and humidity in critical areas under strong disturbances such as instantaneous thermal shock to equipment and seasonal climate changes, without exceeding the absolute red line of 34℃, while gradually approaching the comfort target of 26℃ / 85%, thus meeting the dual needs of the health of downhole workers and the stable operation of equipment.

[0128] For example, the initial real-time wet-bulb black-bulb temperature index at a mine face is 32℃ (compliant but not comfortable). Based on the partial-order grid control strategy, a target grid path with a safety margin M_s=2℃ and a comfort approximation efficiency E_c=0.8℃ / step is selected through a grid path priority evaluation algorithm. The initial control command corresponds to the grid bottom strategy: damper opening 50% (PWM duty cycle 50%), cooling power 70kW (compressor speed 1500r / min), and fan frequency 40Hz. After execution, the real-time wet-bulb black-bulb temperature index is 31℃. According to the closed-loop feedback adjustment rule, the step size Δs=0.5×(26-31)=2 steps (k=0.5, fitted from historical data) is calculated. The strategy moves 2 strategy steps along the target grid path, and the new strategy is damper opening 60% (duty cycle 60%), cooling power 85kW (speed 1800r / min), and fan frequency 45Hz. After execution, the real-time wet-bulb black-bulb temperature index dropped to 28℃. After three more adjustments according to the rules, the comfort target of 26℃ was reached. During this period, a sudden increase in the intensity of a local heat source caused the real-time wet-bulb black-bulb temperature index to briefly rise to 34.2℃. The system immediately reverted to the previous safety grid layer strategy (wet-bulb black-bulb temperature index of 33.5℃). After the heat source intensity recovered, the system continued to approach the comfort target, ultimately achieving stable control of temperature and humidity in the key areas.

[0129] Example 2:

[0130] This embodiment, based on Embodiment 1, provides a system for constructing and dynamically controlling an airflow-thermal-humidity coupling model, such as... Figure 3 As shown, it includes:

[0131] Data fusion and mesh generation module: used to collect multi-source temperature and humidity sensor data and wind network topology, fuse the multi-source temperature and humidity sensor data and wind network topology, generate and output thermal and humidity disturbance phase space mesh;

[0132] Compliance matte generation module: Based on the thermal and moisture perturbation phase space grid, the compliance evolution path is deduced through the thermal and moisture conduction matte generator, the compliance of the evolution path is verified, and the thermal and moisture compliance matte is output.

[0133] The partial order lattice generation module for control strategies is used to determine the control objectives of multiple chambers under the constraints of thermal and humidity compliance matroid, generate and optimize the candidate set of control strategies, construct the partial order relationship between control strategies, and generate the partial order lattice of control strategies through the partial order lattice optimizer for multi-chamber conflict resolution.

[0134] Dynamic closed-loop control module: Based on the partial order grid of the control strategy, it uses multiple actuators guided by the grid path to select the target grid path, generate and update control instructions, coordinate the actions of multiple actuators, execute dynamic closed-loop control instructions, and stabilize the temperature and humidity of key areas within the compliant and comfortable range.

Claims

1. A method for constructing and dynamically controlling a coupled airflow-thermal-humidity model, characterized in that, The method includes: Step S10: Collect multi-source temperature and humidity sensor data and wind network topology, fuse the multi-source temperature and humidity sensor data and wind network topology to generate and output a thermal and humidity disturbance phase space grid; the thermal and humidity disturbance phase space grid is a three-dimensional tensor structure, the tensor dimensions are spatial node dimension, time step dimension and disturbance dimension; the spatial node dimension includes the location of all key areas that need to be monitored in the mine, the time step length is divided based on the seasonal periodic fluctuation law of the mine surface climate, and the disturbance dimension elements are temperature, humidity, wind speed and heat source intensity; Step S20: Based on the thermal and moisture perturbation phase space grid, a compliance evolution path is deduced through a thermal and moisture conduction matte generator, and a thermal and moisture compliance matte is output. The thermal and moisture conduction matte generator specifically includes: extracting thermal and moisture state parameters of each spatial node at different time steps from the thermal and moisture perturbation phase space grid; establishing thermal and moisture state transition rules based on the thermal and moisture state parameters; generating future thermal and moisture state evolution paths using a breadth-first search strategy, starting from the initial thermal and moisture state; verifying the compliance of each generated thermal and moisture state evolution path to obtain a compliance evolution path; and constructing a thermal and moisture compliance matte based on the compliance evolution path using a matte construction algorithm. Step S30: Under the constraints of thermal and humidity compliance matte, a partial-order lattice of control strategies is generated through a partial-order lattice optimizer for multi-chamber conflict resolution. Specifically, the partial-order lattice optimizer for multi-chamber conflict resolution includes: determining the thermal and humidity control objectives for multiple chambers based on the thermal and humidity compliance matte; generating a candidate set of control strategies, including damper opening, cooling power, and fan frequency, and ensuring that the candidate set satisfies the thermal and humidity compliance matte constraints; and performing multi-chamber conflict identification on the candidate set of control strategies. Conflict identification involves substituting each control strategy in the candidate set into the multi-chamber thermal and humidity network model to calculate the wet-bulb black-bulb temperature index after implementing the strategy in each chamber, and comparing the wet-bulb temperature index after implementing the strategy in each chamber. The target set of black sphere temperature index and chamber wet sphere black sphere temperature index is used to count the number of chambers exceeding the target threshold. If the number of exceeding the threshold is greater than 0, the strategy is marked as a conflicting strategy; otherwise, it is marked as a non-conflicting strategy, resulting in a conflicting strategy subset and a non-conflicting strategy subset. The conflicting strategy subset is optimized using a multi-chamber conflict resolution algorithm to obtain a resolved strategy subset. The non-conflicting strategy subset and the resolved strategy subset are merged to obtain a compliant control strategy set. Based on the compliant control strategy set and the partial order relation set, a control strategy partial order lattice is constructed. The partial order relation is defined as strategy A having Pareto advantage over strategy B. The partial order relation set is obtained by pairwise comparison of all strategies in the compliant control strategy set. Step S40: Based on the partial order grid of the control strategy, the multi-actuator guided by the grid path executes the dynamic closed-loop control command to stabilize the temperature and humidity in the key areas of the mine within the compliant and comfortable range; the multi-actuator includes damper actuator, refrigeration unit actuator and fan frequency converter actuator.

2. The method for constructing and dynamically controlling a wind-flow-thermal-humidity coupling model according to claim 1, characterized in that, The process of fusing multi-source temperature and humidity sensor data with wind network topology includes: Spatiotemporal alignment processing is performed on multi-source temperature and humidity sensor data to obtain spatiotemporal aligned data; The spatiotemporal weights are calculated, which include time weights and spatial weights. The time weights are determined based on the time difference between the data acquisition time and the current time and the seasonal cycle variation. The spatial weights are determined based on the straight-line distance between the sensor and the spatial node and the wind resistance coefficient in the wind network topology. The temperature, humidity, wind speed, and heat source intensity in the spatiotemporally aligned data are multiplied by their corresponding time weights and spatial weights to obtain weighted temperature data, weighted humidity data, weighted wind speed data, and weighted heat source intensity data. Weighted temperature data, weighted humidity data, weighted wind speed data, and weighted heat source intensity data are used as quadruples of perturbation dimensions to form a thermal and humid perturbation phase space grid.

3. The method for constructing and dynamically controlling a wind-flow-thermal-humidity coupling model according to claim 1, characterized in that, The matroid construction algorithm includes: Define the base set as the set of all compliance evolution paths that have undergone compliance verification; Define the family of independent sets as a subset of a basic set that satisfies the matroid axioms, which include the heritability axiom and the commutativity axiom; The matroid basis is determined to be the independent set with the largest number of elements in the family of independent sets; Iterate through all elements in the family of independent sets and verify whether they satisfy the axioms of inheritance and commutativity. The structure of the thermo-humidity compliant matte consists of a base set, a family of independent sets, and a matte base.

4. The method for constructing and dynamically controlling a wind-flow-thermal-humidity coupling model according to claim 1, characterized in that, The constructed control strategy partial order lattice includes: Each control strategy in the set of compliant control strategies is used as an element of the grid, and the element attributes include damper opening, cooling power, and fan frequency; Define lattice operations, which include intersection and union operations. The intersection operation is to take the minimum value of the element attributes in two control strategies, and the union operation is to take the maximum value of the element attributes in two control strategies. The cell base is determined as the minimum result of the intersection operation among all strategies, and the cell top is determined as the maximum result of the union operation among all strategies.

5. The method for constructing and dynamically controlling a wind-flow-thermal-humidity coupling model according to claim 1, characterized in that, The grid-path guided multi-executor includes: The target lattice path is selected from the partial order lattice of the control strategy by the lattice path priority evaluation algorithm. The target lattice path must simultaneously satisfy the safety margin and the comfort approximation efficiency constraints. Initial control commands are generated based on the target grid path, and the parameters in the grid bottom strategy are converted into control signals for damper actuators, chiller unit actuators, and fan frequency converter actuators. After sending the initial control command, collect temperature and humidity monitoring data of key areas and calculate the real-time wet-bulb black-bulb temperature index; Based on the real-time wet-bulb black bulb temperature index, the control instructions are updated through closed-loop feedback adjustment rules, including maintaining the comfort range, optimizing the compliance range, and emergency retreat rules for exceeding limits. Avoid action conflicts through collaborative timing control; Built-in fault tolerance mechanism to reselect alternative grid paths in case of equipment failure or sensor malfunction.

6. The method for constructing and dynamically controlling a wind-flow-thermal-humidity coupling model according to claim 5, characterized in that, The closed-loop feedback adjustment rules include: Rule 1: If the real-time wet-bulb temperature index and relative humidity are within the comfort range, maintain the current control strategy and generate a maintenance command. The command content is the maintenance parameter of the current actuator control signal. Rule 2: If the real-time wet-bulb black ball temperature index is in the compliant but not comfortable range, then move along the target grid path towards the aggressive strategy. Rule 3: If the real-time wet-bulb black sphere temperature index exceeds 34°C, a rollback command is immediately generated to roll back to the previous safe cell layer strategy on the target cell path, and the reason for exceeding the limit is recorded.

7. A system for constructing and dynamically controlling a wind-flow-thermal-humidity coupling model, used to implement the wind-flow-thermal-humidity coupling model construction and dynamic control method according to any one of claims 1-6, characterized in that, The system includes: Data fusion and mesh generation module: used to collect multi-source temperature and humidity sensor data and wind network topology, fuse the multi-source temperature and humidity sensor data and wind network topology, generate and output thermal and humidity disturbance phase space mesh; Compliance matte generation module: Based on the thermal and moisture perturbation phase space grid, the compliance evolution path is deduced through the thermal and moisture conduction matte generator, the compliance of the evolution path is verified, and the thermal and moisture compliance matte is output. The partial order lattice generation module for control strategies is used to determine the control objectives of multiple chambers under the constraints of thermal and humidity compliance matroid, generate and optimize the candidate set of control strategies, construct the partial order relationship between control strategies, and generate the partial order lattice of control strategies through the partial order lattice optimizer for multi-chamber conflict resolution. Dynamic closed-loop control module: Based on the partial order grid of the control strategy, it uses multiple actuators guided by the grid path to select the target grid path, generate and update control instructions, coordinate the actions of multiple actuators, execute dynamic closed-loop control instructions, and stabilize the temperature and humidity of key areas within the compliant and comfortable range.

Images