Automatic temperature control demisting monitoring camera system for coal mine field

Abstract

The invention discloses a self-temperature-control demisting monitoring camera system for a coal mine field, particularly relates to the field of coal mine tunnel monitoring, and is used for solving the problems that the imaging quality is interfered by atomization and the demisting response efficiency is low. According to the method, local atomization risks of monitoring points are identified step by step, environment relevance between camera devices is established, demisting linkage is carried out by utilizing real-time triggering conditions, and incremental learning type correction is carried out on energy consumption and environment data after the demisting process, so that the demisting efficiency is improved. The self-adaptive, cooperative and efficient demisting function of the multi-point monitoring device in the coal mine tunnel is achieved, the local atomization degree can be accurately judged, the adjacent camera devices can be rapidly triggered to execute preheating demisting operation, and energy consumption and environment change information can be collected in the demisting process; and long-term stable monitoring and energy utilization optimization in a complex roadway environment are achieved by dynamically correcting an environment correlation model and a demisting power distribution strategy.

Description

Technical Field

[0001] The present invention relates to the field of coal mine tunnel monitoring, and more specifically, to a self-controlled temperature and demisting monitoring camera system for coal mines. Background Art

[0002] The temperature and humidity in coal mine tunnels change frequently, and surveillance cameras are prone to fogging in this environment, resulting in blurred images and affecting the accuracy of safety inspections. Usually, multi-point distributed monitoring devices only rely on the independent defogging operation of a single camera, and it is difficult to link other devices in the same or adjacent tunnels for synchronous protection in the first time, resulting in defogging lag and energy waste.

[0003] Most existing methods only start local defog after a certain device detects fogging, and do not consider the fine control of defog power. There is also a lack of comprehensive analysis of environmental correlation, which results in other monitoring nodes in the same ventilation environment or affected by similar external factors being unable to preheat and defog in advance, resulting in a decline in overall monitoring quality. A distributed predictive control strategy based on historical data and real-time dynamic environmental correlation is urgently needed. After a single point trigger, nodes with potential fogging risks in the same environment lanes can be calculated in a timely manner and advanced defog commands can be issued. At the same time, the defog power of different monitoring devices can be adjusted to achieve coordinated defogging and energy saving and efficiency improvement in lane monitoring.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a self-controlled temperature-controlling defogger monitoring camera system for coal mines, which gradually identifies the local fogging risks of the monitoring points, establishes the environmental correlation between the camera devices, uses real-time trigger conditions for defogger linkage, and performs incremental learning corrections to the energy consumption and environmental data after the defogger process. It realizes the adaptive, collaborative and efficient defogger function of multi-point monitoring devices in coal mine tunnels, which can not only accurately determine the local fogging degree and quickly trigger adjacent camera devices to perform preheating defogger operations, but also collect energy consumption and environmental change information during the defogger process, so as to dynamically correct the environmental correlation model and the defogger power allocation strategy, so as to achieve long-term stable monitoring and energy utilization optimization in complex tunnel environments, so as to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A self-controlled temperature and defogging monitoring camera system for a coal mine, comprising: a fogging assessment module, a correlation analysis module, an instruction generation module, a temperature control and defogging module and a model optimization module;

[0008] Fog assessment module: The monitoring camera device collects environmental data and picture feature information in real time, calculates and generates local fog warning values ​​by analyzing environmental distribution changes and visual blur trends, and transmits the local fog warning values ​​and related environmental data to the correlation analysis module;

[0009] Association analysis module: calculates the environmental similarity between the camera devices through a dynamic association model and identifies the area groups with high environmental similarity, and transmits the dynamic association matrix and area group data to the instruction generation module;

[0010] Instruction generation module: When any camera device triggers a defog signal, the target device affected by the association is identified based on the dynamic association matrix, the preheating defog power of each target device is calculated and the corresponding preheating defog instruction is generated, and the preheating defog instruction is transmitted to the temperature control defog module;

[0011] Temperature control and defogging module: After the target camera receives the preheating and defogging command, the temperature control module is started to dynamically adjust the power according to the real-time environmental status to perform the defogging operation, while real-time monitoring and feedback of energy consumption data and environmental status information to the model optimization module;

[0012] Model optimization module: By analyzing the data sent back by all cameras and combining incremental learning methods to dynamically correct the environmental association model and defogger power allocation strategy, the accuracy and energy efficiency of multi-point linkage defogger are optimized.

[0013] In a preferred embodiment, the atomization assessment module includes the following:

[0014] 1) The heat and humidity diffusion index is designed to evaluate the rate of change and complexity of the temperature and humidity distribution in the environment; based on the sensors installed in the camera devices in various places in the tunnel, the temperature and humidity data are collected in real time to generate the distribution grid G ​​of the monitoring point environment. i (x, y), where i represents the identification number of the camera device, T i (x,y) and H i (x, y) are the temperature and humidity values ​​at the grid points respectively; calculate the spatial gradient of local temperature and humidity: By taking the square root of the sum of the squares of the gradients, we can get the ambient diffusion field strength S i (x,y): Finally, the diffusion field intensity of the monitoring point grid area is integrated to generate the hot and humid flow field diffusion index THFD i : Where dA represents the area of ​​each cell in the grid region.

[0015] In a preferred embodiment, 2), the edge fuzzy expansion index focuses on the visual performance of fog in the monitoring picture; for the real-time picture obtained by the monitoring camera device, the image is first grayed, and the Canny algorithm is used to extract the edge pixel set ε i , calculate the center of gravity position G of the edge point of the picture i,t : Compare the center of gravity offset ΔG of adjacent frames i :ΔG i =|G i,t+1 -G i,t |; At the same time, the fuzzy characteristics of the gradient of the edge points of the picture are calculated, that is, the edge fuzzy expansion index EBE i : in, is the second-order gradient value of the edge point; ∈ is a small positive number.

[0016] In a preferred embodiment, 3), by quantifying the dynamic difference between the wet heat flow field diffusion index and the edge fuzzy expansion index, the synergistic influence of the environment and the image characteristics is captured to generate the atomization comprehensive coefficient C i ; The comprehensive coefficient of atomization is compared with the set threshold C i,th For comparison: If C i ≥C i,th , a local atomization warning signal is generated; otherwise, the monitoring state is maintained;

[0017] 4) After generating the atomization warning signal, calculate the local atomization warning value W i , calculated by the following formula: W i =max(0,C i -C i,th )·ln(1+C i ), where C i -C i,th Reflects the difference between the comprehensive coefficient of fogging at the monitoring point and the corresponding threshold, which is used to measure the extent to which the fogging risk of the current monitoring point exceeds the threshold; if C i ≤C i,th , the warning value is directly 0, indicating that no response is required.

[0018] In a preferred embodiment, the association analysis module includes the following contents:

[0019] Based on the temperature and humidity grid data uploaded by each camera device, the analysis platform constructs the dynamic correlation matrix R through the following formula ij , indicating the environmental similarity between the cameras: Among them, R ij is the environmental similarity between camera i and camera j; and is the temperature and humidity gradient; G i,j is the intersection area of ​​the grids covered by camera devices i and j; for camera device pairs with significant similarity values ​​in the association matrix, R ij ≥R th Perform step-by-step clustering to form multiple regional groups, each of which represents a group of interrelated camera devices, where R th Represents a similarity threshold; records the identification numbers of the camera devices in each group and their associations.

[0020] In a preferred embodiment, the instruction generation module includes the following contents:

[0021] Each camera device continuously monitors its own environmental status. When its local fogging warning value exceeds the set defogger trigger threshold, it generates a defogger trigger signal and immediately uploads the signal together with the current environmental parameters to the central analysis platform. After receiving the trigger signal, the central analysis platform quickly screens the target device set T with high environmental similarity to the triggered camera device i through the association matrix generated by the association analysis module. i ; The target device is selected based on the following conditions: T i ={j|R ij ≥R th and W j <W j,th}; where j is the camera device identification number; for the selected target device set, combined with the latest environmental parameters of each target device, the central analysis platform calculates its preheating defogging power P j :P j =P base ·(1+ln(1+R ij ·(W i -W i,th ))); where P base Is the basic preheating power; W i -W i,th Reflects the degree to which the risk of atomization of the trigger device exceeds the limit; R ij The response priority of the device with high similarity is amplified; the calculated preheating and defogging power is used to generate a preheating and defogging instruction, and is sent to each target camera device.

[0022] In a preferred embodiment, the temperature control and demisting module includes the following contents:

[0023] After receiving the preheating and defogging command sent by the central analysis platform, the target camera device analyzes the preheating and defogging power and the estimated response time in the command; the camera device internally checks whether the current power status and environmental conditions meet the basic requirements for executing the command. If not, it notifies the central analysis platform through a return signal to make adjustments; based on the preheating and defogging power, the temperature control module is started and dynamic power adjustment is implemented. The adjustment formula is as follows: Among them, P j,t is the temperature control power at time t; ΔT j (t) is the real-time temperature change of the target camera device at time t; Δt is the sampling time interval; during the defogging operation, the target camera device monitors the energy consumption parameter E during the defogging period in real time j,t and environmental status data, and transmit the data back to the central analysis platform at set time intervals.

[0024] In a preferred embodiment, the model optimization module includes the following contents:

[0025] Based on the data sent back by each camera device, the analysis platform calculates its actual energy efficiency ratio R eff,j : Among them, T j,init and T j,final are the initial and final temperature distributions before and after demisting, respectively; t start and t end The start and end time of the defog operation.

[0026] In a preferred embodiment, the model optimization module further includes the following contents:

[0027] The analysis platform combines the environmental grid data transmitted back during the defogging process to correct the similarity matrix of the dynamic association model. The correction formula is: Where α is the environmental adjustment factor; ΔT j and ΔT i is the temperature variation of the camera devices j and i.

[0028] In a preferred embodiment, the model optimization module further includes the following contents:

[0029] The incremental learning method is used to add the latest feedback from each defog operation process into the training set, and the defog trigger judgment and power allocation model are dynamically optimized.

[0030] The technical effects and advantages of the self-controlled temperature defogging monitoring camera system for coal mines of the present invention are as follows:

[0031] The present invention realizes the adaptive, collaborative and efficient defog function of multi-point monitoring devices in coal mine tunnels by gradually identifying the local fogging risks of monitoring points, establishing the environmental correlation between camera devices, utilizing real-time trigger conditions for defog linkage, and performing incremental learning-based corrections to energy consumption and environmental data after the defog process. It can not only accurately determine the local fogging degree and quickly trigger adjacent camera devices to perform preheating defog operations, but also collect energy consumption and environmental change information during the defog process, so as to dynamically correct the environmental correlation model and the defog power allocation strategy, so as to achieve long-term stable monitoring and energy utilization optimization in complex tunnel environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 The present invention is a structural schematic diagram of a self-controlled temperature demisting monitoring camera system for coal mines. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] Embodiment 1: Figure 1 The invention provides a self-controlled temperature defogging monitoring camera system for coal mines, comprising: a fogging evaluation module, a correlation analysis module, an instruction generation module, a temperature control defogging module and a model optimization module;

[0035] Fog assessment module: The monitoring camera device collects environmental data and picture feature information in real time, calculates and generates local fog warning values ​​by analyzing environmental distribution changes and visual blur trends, and transmits the local fog warning values ​​and related environmental data to the correlation analysis module;

[0036] Association analysis module: calculates the environmental similarity between the camera devices through a dynamic association model and identifies the area groups with high environmental similarity, and transmits the dynamic association matrix and area group data to the instruction generation module;

[0037] Instruction generation module: When any camera device triggers a defog signal, the target device affected by the association is identified based on the dynamic association matrix, the preheating defog power of each target device is calculated and the corresponding preheating defog instruction is generated, and the preheating defog instruction is transmitted to the temperature control defog module;

[0038] Temperature control and defogging module: After the target camera receives the preheating and defogging command, the temperature control module is started to dynamically adjust the power according to the real-time environmental status to perform the defogging operation, while real-time monitoring and feedback of energy consumption data and environmental status information to the model optimization module;

[0039] Model optimization module: By analyzing the data sent back by all cameras and combining incremental learning methods to dynamically correct the environmental association model and defogger power allocation strategy, the accuracy and energy efficiency of multi-point linkage defogger are optimized.

[0040] The environmental conditions in coal mine tunnels are complex. The temperature and humidity distribution and dynamic changes at each monitoring point have a significant impact on the generation of fogging risk, and the change in the clarity of the monitoring image directly reflects the interference of fogging on the imaging quality. By independently collecting and analyzing environmental data and image feature information for each monitoring point, the fogging risk level of the local area can be accurately identified, providing a high-precision initial judgment basis for subsequent defog operations, and realizing the real-time and targeted nature of fogging risk assessment.

[0041] The fogging assessment module includes the following:

[0042] 1) The humid heat flow field diffusion index is designed to evaluate the rate of change and complexity of the temperature and humidity distribution in the environment. The formation of fog usually stems from the local inhomogeneity of temperature and humidity. Especially when the humidity rises rapidly and the temperature drops, it is more likely to trigger the condensation effect, causing the lens to fog. By calculating the spatial gradient of temperature and humidity, it is possible to quantify the diffusion field intensity in the local area and identify potential high fogging risk areas. This index provides the system with a dynamic description of the physical conditions of the environment.

[0043] Based on the sensors installed in the camera devices in various places in the lane, the temperature and humidity data are collected in real time to generate the distribution grid G ​​of the monitoring point environment. i (x, y), where i represents the identification number of the camera device, T i (x,y) and H i (x, y) are the temperature and humidity values ​​at the grid point respectively. For this grid, calculate the spatial gradient of local temperature and humidity:

[0044] By taking the square root of the sum of the squares of the gradients, we can get the ambient diffusion field strength S i (x,y):

[0045] Finally, the diffusion field intensity of the monitoring point grid area is integrated to generate the hot and humid flow field diffusion index THFD i : Where dA represents the area of ​​each unit in the grid area, which is used to perform area-weighted calculation of the diffusion field intensity.

[0046] The wet heat flow field diffusion index is used to quantify the rate of change and complexity of the temperature and humidity distribution in the environment, reflecting the spatial diffusion characteristics of the temperature gradient and humidity gradient in the local area. The larger the value, the more uneven the distribution of temperature and humidity near the monitoring point, the higher the heat conduction and moisture diffusion rate in the local environment, indicating that the risk of fogging has increased significantly; the smaller the value, the more uniform the temperature and humidity distribution, the more stable the environmental changes, and the lower the risk of fogging.

[0047] 2) Edge Blur Extension Index focuses on the visual performance of fog in the monitoring screen. When fog begins to condense, the image clarity of the lens will decrease, which is manifested as blurred edge features, reduced sharpness, and center of gravity shift. The Edge Blur Extension Index captures the changes in optical properties caused by fog by analyzing the blur extension trend of the edge of the picture.

[0048] For the real-time images obtained by the surveillance camera, the image is first grayed out and the Canny algorithm is used to extract the edge pixel set ε i , calculate the center of gravity position G of the edge point of the picture i,t :

[0049] Compare the center of gravity offset ΔG of adjacent frames i :ΔG i =|G i,t+1 -G i,t |;

[0050] At the same time, the fuzzy characteristics of the gradient of the edge points of the picture are calculated, that is, the edge fuzzy expansion index EBE i : in, It is the second-order gradient value of the edge point, which is used to quantify the degree of edge blur; ∈ is a small positive number to avoid the denominator being zero.

[0051] The edge blur extension index reflects the quantitative characteristics of the decrease in edge clarity and the expansion of blur caused by fog in the monitoring picture. The larger the value, the more significant the blur range and blur degree of the edge in the picture, indicating that the interference of fog on the visual imaging quality is more serious; the smaller the value, the clearer the edge features of the picture are, the fog has not had a significant impact on the monitoring picture, and the imaging quality is stable.

[0052] 3) By quantifying the dynamic difference between the diffusion index of the moist heat flow field and the edge fuzzy expansion index, the synergistic influence of the environment and the picture characteristics is captured to generate the comprehensive coefficient of fogging C iFirst, the significance of the environment and picture features is significantly amplified through the squared difference term, and the exponential function is introduced to moderately smooth the impact of the picture features, thereby improving the sensitivity to slight changes. Secondly, by adjusting the inverse tangent of the exponential product, nonlinear enhancement is achieved in high-risk scenarios, while effectively controlling excessive deviations caused by outliers. The calculation of the comprehensive coefficient can not only accurately reflect the interaction between physical environmental conditions and visual picture features, but also maintain the smoothness and robustness of the results under different fogging risk levels, providing an accurate basis for the generation of fogging warning signals. The calculation formula is:

[0053] Part I By calculating the square of the difference between the two exponents, the significance of the environment and the picture features is amplified, and the exp(-EBE i ) controls the effect of the picture index on the overall coefficient.

[0054] Part 2 The inverse tangent function is used to perform nonlinear adjustment on the product of exponents to enhance the comprehensive response capability in high-risk scenarios.

[0055] The comprehensive coefficient of atomization is compared with the set threshold C i,th For comparison:

[0056] If C i ≥C i,th , a local atomization warning signal is generated; otherwise, the monitoring state is maintained.

[0057] This judgment comprehensively determines the atomization risk based on the environmental conditions and image characteristics of the monitoring point.

[0058] 4) After generating the atomization warning signal, further calculate the local atomization warning value W i , for example, calculated by the following formula: W i =max(0,C i -C i,th )·ln(1+C i );

[0059] C i -C i,th Reflects the difference between the comprehensive coefficient of fogging at the monitoring point and the corresponding threshold, which is used to measure the extent to which the fogging risk of the current monitoring point exceeds the threshold; if C i ≤C i,th , the warning value is directly 0, indicating that no further response is required.

[0060] ln(1+C i) performs logarithmic enhancement on the comprehensive fog coefficient, so that higher comprehensive fog coefficient values ​​obtain higher weights when calculating local fog warning values, reflecting the priority response of high-risk scenarios.

[0061] max(0,·) ensures that the warning value is non-negative and avoids meaningless negative values ​​when the threshold is higher than the comprehensive coefficient.

[0062] Through comprehensive analysis of the environmental data and image features of the monitoring points, quantitative assessment and dynamic response to fogging risks are achieved. It can accurately distinguish abnormal changes in the local environment and timely capture the degree of interference of fogging on imaging quality, providing real-time support for the stable operation of monitoring equipment while reducing misjudgments and unnecessary operational interventions.

[0063] In the fog assessment module, each camera device generates a local fog warning value based on its environmental data and image features, which is an independent assessment of the fog risk of its respective environment. Since the environments of the cameras in the coal mine tunnels are somewhat correlated, the fog risk in the area where a certain device is located may affect other devices through mechanisms such as air flow fields and temperature and humidity diffusion. The correlation analysis module summarizes the fog warning values ​​and environmental parameters of each camera device through the central analysis platform, constructs a dynamic correlation model, calculates the environmental similarity between each camera device, and identifies regional groups that can have mutual influence, providing support for subsequent linkage defogging decisions.

[0064] The association analysis module includes the following:

[0065] Each camera uploads the local fog warning value and the corresponding temperature and humidity grid data and image feature analysis value (edge ​​blur expansion index) to the central analysis platform through the data network. The central analysis platform performs time synchronization and spatial alignment on the uploaded data to ensure that the input data of each camera is formatted consistently and to eliminate invalid or erroneous data.

[0066] Based on the temperature and humidity grid data uploaded by each camera device, the analysis platform constructs the dynamic correlation matrix R through the following formula ij , indicating the environmental similarity between the cameras: Among them, R ij is the environmental similarity between camera i and camera j; and is the temperature and humidity gradient; G i,j It is the intersection area of ​​the grids covered by cameras i and j.

[0067] For the camera device pairs with significant similarity values ​​in the correlation matrix, R ij ≥R thPerform step-by-step clustering to form multiple regional groups, each of which represents a group of interrelated camera devices, where R th Represents the similarity threshold. Record the identification numbers of the cameras in each group and their associations for subsequent linkage decisions.

[0068] Based on the statistical data of the historical association matrix and the real-time uploaded environmental parameters, the regional groups are dynamically updated. The analysis platform recursively optimizes the association matrix after each new data input, so that each regional group can accurately reflect the dynamic changes of the current environmental conditions.

[0069] Through centralized processing of fog warning values ​​and environmental data uploaded by cameras, a dynamic correlation model between cameras was constructed, and regional groups with high environmental similarity were identified. Accurate quantification of environmental correlation in coal mine tunnels was achieved, providing a scientific basis for multi-point linkage fog response strategies.

[0070] In the fog assessment module, each camera device generates a local fog warning value to quantify its own fog risk. In the association analysis module, the central analysis platform calculates the environmental similarity between the cameras through a dynamic association model and identifies the interrelated regional groups. Based on these associated data, the instruction generation module needs to determine the defogging conditions that occur in real time. When any camera device triggers the defogging requirement, it quickly identifies other cameras in the group that are affected by the association, determines potential preheating defogging targets, and issues preheating defogging instructions to ensure the timeliness and accuracy of the response.

[0071] The instruction generation module includes the following:

[0072] Each camera device continuously monitors its own environmental status. When its local fogging warning value exceeds the set defogger trigger threshold, it generates a defogger trigger signal and immediately uploads the signal together with the current environmental parameters (including the latest temperature and humidity distribution data and edge blur expansion index) to the central analysis platform.

[0073] After receiving the trigger signal, the central analysis platform quickly screens the target device set T with high environmental similarity to the triggered camera device i through the association matrix generated by the association analysis module i The target device selection is based on the following criteria: i ={j|R ij ≥R th and W j <W j,th}; where j is the camera device identification number. The condition ensures that the selected target device has not triggered defogger, but has a high probability of being affected by the triggering device.

[0074] For the selected target device set, combined with the latest environmental parameters of each target device, the central analysis platform calculates its preheating defogging power P j :P j =P base ·(1+ln(1+R ij ·(W i -W i,th ))); where P base Is the basic preheating power; W i -W i,th Reflects the degree to which the risk of atomization of the trigger device exceeds the limit; R ij Amplify the response priority of high similarity devices.

[0075] The calculated preheating and defogging power is used to generate a refined preheating and defogging instruction, and is sent to each target camera device.

[0076] The central analysis platform transmits the preheating and demisting instructions to the target device set. The instructions contain the preheating and demisting power and expected response time information of the target device, ensuring that each target device can adjust the temperature control module as needed and perform the preheating and demisting operation in a timely manner.

[0077] By combining the dynamic association model with real-time trigger judgment, the target camera device affected by the association can be quickly identified after the defogging condition occurs, and an accurate preheating defogging instruction can be issued. This processing logic ensures the timeliness and relevance of the defogging operation, effectively improves the collaborative efficiency of multi-point defogging, and avoids unnecessary power waste.

[0078] In the instruction generation module, the central analysis platform identifies the target camera device that may be affected by fogging by performing correlation analysis on the camera device that triggers the defog signal, and generates a preheating defog instruction based on the real-time environmental data. The task of the temperature control defog module is to start the temperature control module for preheating defog according to the preheating defog power after the target camera device receives the instruction, and at the same time, transmit the energy consumption data and environmental parameters generated during the defog process back to the central analysis platform for subsequent correction and optimization of the dynamic correlation model.

[0079] The temperature control and demisting module includes the following:

[0080] After the target camera receives the preheating and defogging command sent by the central analysis platform, it analyzes the preheating and defogging power and the estimated response time in the command. The preheating and defogging power represents the initial power setting value of the temperature control module, and the estimated response time is used to control the timeliness of defogging. The camera internally checks whether the current power status and environmental conditions meet the basic requirements for executing the command. If not, it notifies the central analysis platform through a return signal to make adjustments.

[0081] According to the preheating and demisting power, the temperature control module is started and dynamic power adjustment is implemented. The adjustment formula is as follows: Among them, P j,t is the temperature control power at time t; ΔT j (t) is the real-time temperature change of the target camera device at time t; Δt is the sampling time interval, which is used to standardize the adjustment amplitude. Dynamic power adjustment optimizes the defogging process according to the real-time temperature change to avoid excessive energy consumption or insufficient defogging effect.

[0082] During the defogging operation, the target camera device monitors the energy consumption parameter E during the defogging period in real time. j,t and environmental status data (including temperature and humidity grid G j (x, y) and atomization change index), and transmit the data back to the central analysis platform at the set time interval. The format of the returned data is: j ={t,P j,t ,E j,t ,G j (x,y)}; where t is the data collection time, P j,t is the real-time power, E j,t is the energy consumption value, G j (x,y) is the environmental distribution data.

[0083] When the defog power reaches a stable state and the fog index is lower than the preset clarity threshold, the target camera device confirms that the defog is complete and sends an end signal to the central analysis platform, along with evaluation data on the final defog effect, including total energy consumption and environmental recovery indicators, for optimization of subsequent steps.

[0084] The target camera device accurately responds to the preheating and defogging instructions and provides real-time data feedback, which enables dynamic power optimization of the temperature control module and effective removal of fogging interference. The monitoring and feedback of energy consumption and environmental conditions during the defogging process provide reliable basic data for the continuous optimization of multi-point collaborative defogging strategies.

[0085] In the temperature control and demisting module, after the target camera device receives the preheating demisting command, it completes the temperature control and demisting operation and transmits back the energy consumption data and environmental status information in real time. These feedback data not only reflect the actual effect of the current demisting operation, but also provide a basis for the subsequent optimization of the multi-point linkage self-controlled temperature demisting strategy. The core task of the model optimization module is that the central analysis platform dynamically corrects the environmental association model and the demisting power allocation strategy based on the feedback data of all cameras and combines the incremental learning method, so as to achieve more accurate demisting control and energy efficiency optimization in the complex and changeable coal mine tunnel environment.

[0086] The model optimization module includes the following:

[0087] Based on the data sent back by each camera device, the analysis platform calculates its actual energy efficiency ratio R eff,j : Among them, T j,init and T j,final are the initial and final temperature distributions before and after demisting, respectively; t start and t end The start and end time of the defog operation.

[0088] The power allocation strategy of each camera device is dynamically adjusted according to the energy efficiency ratio to improve the overall defogging efficiency.

[0089] The analysis platform combines the environmental grid data transmitted during the defogging process to correct the similarity matrix of the dynamic association model. The correction formula is: Where α is the environmental adjustment factor; ΔT j and ΔT i is the temperature variation of the camera devices j and i.

[0090] The corrected correlation matrix reflects the actual environmental correlation characteristics more accurately and provides support for subsequent similarity calculations.

[0091] The incremental learning method is used to add the latest feedback from each defog operation process into the training set, and the defog trigger judgment and power allocation model are dynamically optimized.

[0092] The incremental learning model is updated by taking the energy efficiency ratio R transmitted during each defogging operation into account. eff,j , Environmental distribution data G j (x,y), real-time power P j,t , Temperature change range ΔT before and after defogging j ,ΔT i , and data such as defogging effect indicators are dynamically added to the training set to construct a time series feature stream. During the training process, based on the pattern characteristics of historical data, the newly added data is used as a new feature point. Through online optimization training, the model is prevented from forgetting the characteristics of early data, while ensuring that the model can capture the latest environmental changes. Finally, the model outputs the optimized defogging trigger threshold W′ i,th and preheating power parameter P′ j , and adjust the similarity judgment rules of the associated model to adapt to the dynamic changes under complex environmental conditions. Through continuous iteration, the model can gradually improve the response efficiency and accuracy of the multi-point linkage defogging strategy, and accurately adapt to the complex environmental conditions in the coal mine tunnels.

[0093] By integrating and analyzing the feedback data, the environmental correlation model and the defogging power allocation strategy are dynamically corrected to improve the accuracy and energy efficiency of the linkage defogging. The introduction of the incremental learning method enables the model to have continuous optimization capabilities and achieve long-term stable defogging control effects in the complex and changeable coal mine tunnel environment.

[0094] The present invention realizes the adaptive, collaborative and efficient defog function of multi-point monitoring devices in coal mine tunnels by gradually identifying the local fogging risks of monitoring points, establishing the environmental correlation between camera devices, utilizing real-time trigger conditions for defog linkage, and performing incremental learning-based corrections to energy consumption and environmental data after the defog process. It can not only accurately determine the local fogging degree and quickly trigger adjacent camera devices to perform preheating defog operations, but also collect energy consumption and environmental change information during the defog process, so as to dynamically correct the environmental correlation model and the defog power allocation strategy, so as to achieve long-term stable monitoring and energy utilization optimization in complex tunnel environments.

[0095] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0096] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0097] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including a..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0098] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A self-controlled temperature and demisting monitoring camera system for coal mines, characterized in that: include: Atomization evaluation module, correlation analysis module, instruction generation module, temperature control and defogging module, and model optimization module; Fog assessment module: The monitoring camera device collects environmental data and picture feature information in real time, calculates and generates local fog warning values ​​by analyzing environmental distribution changes and visual blur trends, and transmits the local fog warning values ​​and related environmental data to the correlation analysis module; Association analysis module: calculates the environmental similarity between the camera devices through a dynamic association model and identifies the area groups with high environmental similarity, and transmits the dynamic association matrix and area group data to the instruction generation module; Instruction generation module: When any camera device triggers a defog signal, the target device affected by the association is identified based on the dynamic association matrix, the preheating defog power of each target device is calculated and the corresponding preheating defog instruction is generated, and the preheating defog instruction is transmitted to the temperature control defog module; Temperature control and defogging module: After the target camera receives the preheating and defogging command, the temperature control module is started to dynamically adjust the power according to the real-time environmental status to perform the defogging operation, while real-time monitoring and feedback of energy consumption data and environmental status information to the model optimization module; Model optimization module: By analyzing the data sent back by all cameras and combining incremental learning methods to dynamically correct the environmental association model and defogger power allocation strategy, the accuracy and energy efficiency of multi-point linkage defogger are optimized.

2. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 1 is characterized by: The fogging assessment module includes the following: 1) The heat and humidity diffusion index is designed to evaluate the rate of change and complexity of the temperature and humidity distribution in the environment; based on the sensors installed in the camera devices in various places in the tunnel, the temperature and humidity data are collected in real time to generate the distribution grid G ​​of the monitoring point environment. i (x, y), where i represents the identification number of the camera device, T i (x,y) and H i (x, y) are the temperature and humidity values ​​at the grid points respectively; calculate the spatial gradient of local temperature and humidity: By taking the square root of the sum of the squares of the gradients, we can get the ambient diffusion field strength S i (x,y): Finally, the diffusion field intensity of the monitoring point grid area is integrated to generate the hot and humid flow field diffusion index THFD i : Where dA represents the area of ​​each cell in the grid region.

3. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 2 is characterized by: 2) The edge blur expansion index focuses on the visual performance of fog in the monitoring screen; for the real-time screen obtained by the monitoring camera device, the image is first grayed out, and the Canny algorithm is used to extract the edge pixel set ε i , calculate the center of gravity position G of the edge point of the picture i,t : Compare the center of gravity offset ΔG of adjacent frames i :ΔG i =|G i,t+1 -G i,t |; At the same time, the fuzzy characteristics of the gradient of the edge points of the picture are calculated, that is, the edge fuzzy expansion index EBE i : in, (x j ,y j ) is the second-order gradient value of the edge point; ∈ is a small positive number.

4. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 3 is characterized by: 3) By quantifying the dynamic difference between the diffusion index of the moist heat flow field and the edge fuzzy expansion index, the synergistic influence of the environment and the picture characteristics is captured to generate the comprehensive coefficient of fogging C i ; The comprehensive coefficient of atomization is compared with the set threshold C i,th For comparison: If C i ≥C i,th , a local atomization warning signal is generated; Otherwise, keep monitoring; 4) After generating the atomization warning signal, calculate the local atomization warning value W i , calculated by the following formula: W i =max(0,C i -C i,th )·ln(1+C i ), where C i -C i,th Reflects the difference between the comprehensive coefficient of fogging at the monitoring point and the corresponding threshold, which is used to measure the extent to which the fogging risk of the current monitoring point exceeds the threshold; if C i ≤C i,th , the warning value is directly 0, indicating that no response is required.

5. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 4 is characterized in that: The association analysis module includes the following: Based on the temperature and humidity grid data uploaded by each camera device, the analysis platform constructs the dynamic correlation matrix R through the following formula ij , indicating the environmental similarity between the cameras: Among them, R ij is the environmental similarity between camera i and camera j; and is the temperature and humidity gradient; G i,j is the intersection area of ​​the grids covered by camera devices i and j; for camera device pairs with significant similarity values ​​in the association matrix, R ij ≥R th Perform step-by-step clustering to form multiple regional groups, each of which represents a group of interrelated camera devices, where R th Represents a similarity threshold; records the identification numbers of the camera devices in each group and their associations.

6. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 5, characterized in that: The instruction generation module includes the following: Each camera device continuously monitors its own environmental status. When its local fogging warning value exceeds the set defogger trigger threshold, it generates a defogger trigger signal and immediately uploads the signal together with the current environmental parameters to the central analysis platform. After receiving the trigger signal, the central analysis platform quickly screens the target device set T with high environmental similarity to the triggered camera device i through the association matrix generated by the association analysis module. i ; The target device is selected based on the following conditions: T i ={j|R ij ≥R th and W j <W j,th }; where j is the camera device identification number; for the selected target device set, combined with the latest environmental parameters of each target device, the central analysis platform calculates its preheating defogging power P j :P j =P base ·(1+ln(1+R ij ·(W i -W i,th ))); where P base Is the basic preheating power; W i -W i,th Reflects the degree to which the risk of atomization of the trigger device exceeds the limit; R ij The response priority of the device with high similarity is amplified; the calculated preheating and defogging power is used to generate a preheating and defogging instruction, and is sent to each target camera device.

7. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 6 is characterized by: The temperature control and demisting module includes the following: After receiving the preheating and defogging command sent by the central analysis platform, the target camera device analyzes the preheating and defogging power and the estimated response time in the command; the camera device internally checks whether the current power status and environmental conditions meet the basic requirements for executing the command. If not, it notifies the central analysis platform through a return signal to make adjustments; based on the preheating and defogging power, the temperature control module is started and dynamic power adjustment is implemented. The adjustment formula is as follows: Among them, P j,t is the temperature control power at time t; ΔT j (t) is the real-time temperature change of the target camera device at time t; Δt is the sampling time interval; during the defogging operation, the target camera device monitors the energy consumption parameter E during the defogging period in real time j,t and environmental status data, and transmit the data back to the central analysis platform at set time intervals.

8. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 7 is characterized in that: The model optimization module includes the following: Based on the data sent back by each camera device, the analysis platform calculates its actual energy efficiency ratio R eff,j : Among them, T j,init and T j,final are the initial and final temperature distributions before and after demisting, respectively; t start and t end The start and end time of the defog operation.

9. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 8, characterized in that: The model optimization module also includes the following: The analysis platform combines the environmental grid data transmitted back during the defogging process to correct the similarity matrix of the dynamic association model. The correction formula is: Where α is the environmental adjustment factor; ΔT j and ΔT i is the temperature variation of the camera devices j and i.

10. The automatic temperature control and demisting monitoring camera system for coal mines according to claim 9, characterized in that: The model optimization module also includes the following: The incremental learning method is used to add the latest feedback from each defog operation process into the training set, and the defog trigger judgment and power allocation model are dynamically optimized.

Images