A smoke exhaust fan control system and method

Abstract

This invention discloses a smoke exhaust fan control system and method, relating to the field of fire protection system technology. It predicts the performance of the smoke exhaust fan through a loss calculation unit, obtains the operating threshold of the smoke exhaust fan through a temperature curve unit, identifies early smoke characteristics through a smoke monitoring unit, and obtains the smoke diffusion rate and the time required to invade preset proportions of different fire zones through a smoke diffusion unit. Then, it constructs a first model and a second model through a temperature change construction unit to determine whether the smoke exhaust fan reaches the temperature threshold under different conditions, enabling different levels of early warning and corresponding measures to be implemented. This minimizes the decrease in smoke exhaust efficiency of the entire fire protection smoke exhaust system caused by temperature-induced performance degradation of the smoke exhaust fan. In other words, it achieves precise control of the smoke exhaust fan through a central control unit, thereby improving the overall smoke exhaust efficiency of the fire protection smoke exhaust system.

Description

Technical Field

[0001] This invention relates to the field of fire protection system technology, and in particular to a smoke exhaust fan control system and method. Background Technology

[0002] In modern buildings, fire smoke extraction systems play a crucial role as an important part of the safety system. They can quickly remove smoke in the event of a fire, providing a valuable time window for personnel evacuation and fire rescue. This system is usually composed of key components such as smoke exhaust outlets, smoke exhaust ducts, smoke exhaust fans, control systems, and fire dampers. In the event of a fire, these components work together to remove smoke from the building.

[0003] In small buildings, a smoke exhaust fan typically consists of one smoke exhaust fan and multiple smoke exhaust ducts. When a smoke detector detects a certain concentration of smoke, the smoke exhaust fan operates to expel the smoke. However, in large buildings, there are multiple fire compartments, each containing a smoke exhaust fan. When a fire occurs in a single fire compartment, the corresponding smoke exhaust fan and the smoke exhaust fans in other fire compartments will start working. However, the smoke exhaust fans generate a lot of heat when they are working, and since the smoke also carries a lot of heat, the operating environment temperature of the smoke exhaust fans will rise, thereby reducing the optimal performance of the smoke exhaust fans, resulting in a decrease in smoke exhaust volume and efficiency, and ultimately affecting the smoke exhaust efficiency of the entire fire smoke exhaust system. Summary of the Invention

[0004] In view of this, this application provides a smoke exhaust fan control system and method to achieve precise control of the smoke exhaust fan and improve the smoke exhaust efficiency of the entire fire smoke exhaust system.

[0005] The first aspect of this application provides a smoke exhaust fan control system, comprising:

[0006] Loss Calculation Unit: Constructs and trains a loss model, and based on the trained loss model, calculates the performance loss of all smoke exhaust fans in each fire compartment to obtain the loss calculation results;

[0007] Temperature curve acquisition unit: acquires the first coordinate curve of the maximum power of the exhaust fan as a function of temperature, cleans the data in the first coordinate curve based on the loss calculation results, and determines the temperature corresponding to the maximum power drop of the exhaust fan based on the cleaned first coordinate curve, and records it as the temperature threshold.

[0008] Temperature change construction unit: In the preset simulation software, a first model is constructed to model the temperature change of the exhaust fan from startup to operation at maximum power in a smoke-free state; and a second model is constructed to model the temperature change of the exhaust fan when it operates at maximum power in a smoke-containing state; the maximum power of the exhaust fan in the first model and the second model are the actual maximum power calculated by the loss model;

[0009] Smoke monitoring unit: Constructs and trains a smoke recognition model, and identifies early smoke characteristics in each fire compartment based on the trained smoke recognition model;

[0010] Smoke diffusion unit: Construct and train a smoke diffusion model, and simulate the diffusion of smoke in each fire compartment based on the trained smoke diffusion model;

[0011] Central control unit: When a fire occurs in any fire compartment, the smoke exhaust fan in the corresponding fire compartment is activated, and the data from the loss calculation unit, the temperature curve acquisition unit, the temperature change construction unit and the smoke diffusion unit are collected, analyzed and processed, and the smoke exhaust fans in the remaining fire compartments are precisely controlled.

[0012] In one possible implementation of the first aspect, constructing and training the loss model includes:

[0013] Construct a set of loss functions for the air volume, air pressure, maximum power, efficiency, and energy consumption of the exhaust fan, specifically:

[0014]

[0015] This represents the percentage of air volume loss. This is the weighting coefficient for air volume. This refers to the actual air volume. For the target air volume, This represents the proportion of wind pressure loss. This is the weighting coefficient for wind pressure. This is the actual wind pressure. For target wind pressure, This represents the percentage of maximum power loss. The weighting factor for maximum power. This is the actual maximum power. For the target maximum power, This represents the percentage loss in conversion efficiency. The weighting coefficient for conversion efficiency. For actual conversion efficiency, To achieve the target conversion efficiency, This represents the proportion of energy loss. This is a weighting factor for energy consumption. Actual energy consumption Target energy consumption;

[0016] Based on the aforementioned set of loss functions, the loss coefficients of the exhaust fan are constructed as follows:

[0017]

[0018] The loss coefficients of the exhaust fan at multiple time scales are obtained, and the results are obtained.

[0019] Based on the obtained results, a coordinate curve showing the change of the loss coefficient with the usage time of the exhaust fan is constructed with the usage time as the horizontal axis and the loss coefficient of the exhaust fan as the vertical axis, and then fitted.

[0020] In one possible implementation of the first aspect, performance loss calculations are performed on all smoke exhaust fans in each fire compartment, and the loss calculation results include:

[0021] The usage time of any smoke exhaust fan is recorded as the first usage time, and the fitted coordinate curve is recorded as the second coordinate curve.

[0022] Based on the fitted second coordinate curve, the loss coefficient at the first usage time node is predicted and denoted as the first loss coefficient.

[0023] Based on the first loss coefficient, the loss calculation results of the corresponding smoke exhaust fan are obtained.

[0024] In one possible implementation of the first aspect, cleaning the data in the first coordinate curve based on the loss calculation result includes:

[0025] Based on the loss calculation results, the actual maximum power of the exhaust fan is obtained;

[0026] In the first coordinate curve, the data point corresponding to the actual maximum power is matched, and the curve data located before the data point is cleaned.

[0027] In one possible implementation of the first aspect, determining the temperature threshold based on the first coordinate curve after cleaning includes:

[0028] In the first coordinate curve, taking the data point corresponding to the actual maximum power as the starting point, the first coordinate curve located after the data point is divided into several equal intervals, and the rate of change of the first coordinate curve in each interval is calculated to obtain the rate of change calculation result.

[0029] Based on the calculation results of the rate of change, the interval corresponding to the largest rate of change is selected and denoted as the first interval;

[0030] The temperature value at the left port of the first interval is recorded as the temperature threshold.

[0031] In one possible implementation of the first aspect, constructing and training the smoke recognition model includes:

[0032] Collect surveillance video data at multiple time scales as a sample set and store it in the database;

[0033] Data augmentation is performed on the sample set in the database to expand the data volume of the sample set;

[0034] Manually label the data-enhanced sample set;

[0035] Based on a manually labeled sample set, a region candidate network is used to generate candidate regions containing smoke.

[0036] A convolutional neural network is used to classify and regress the candidate region, and output the category and location coordinates of the smoke.

[0037] In one possible implementation of the first aspect, constructing and training the smoke diffusion model includes:

[0038] Construct a smoke diffusion model and set initial parameters for the smoke diffusion model, including BIM data of the fire compartment, location of the smoke source, and physical quantities of the smoke source;

[0039] Establish wind force model, diffusion force model and air resistance model and solve them separately. Combine the solution results to obtain external force data.

[0040] Based on the initial parameters and the merged external force data, the smoke diffusion model is solved, and the velocity field, temperature field and density field of the smoke are updated.

[0041] Based on the size of the fire compartment and the boundaries of obstacles, the movement of smoke is constrained, thereby correcting the velocity field, temperature field and density field of the smoke;

[0042] Based on the modified velocity field, temperature field, and density field of the smoke in the smoke diffusion model, the diffusion of smoke is simulated and plotted in real time.

[0043] In one possible implementation of the first aspect, activating the smoke exhaust fan in the corresponding fire compartment when a fire occurs includes:

[0044] Based on the smoke monitoring unit, the type of smoke and the coordinates of its location in any fire compartment can be obtained;

[0045] If the smoke is determined to be fire smoke, the smoke exhaust fan of the corresponding fire compartment is turned on, and the location coordinates of the fire smoke are uploaded to the preset terminal.

[0046] In one possible implementation of the first aspect, precise control of the smoke exhaust fans in the remaining fire compartments includes:

[0047] When there is fire smoke in fire compartment A, the time point at which the fire smoke occurs is recorded as t0.

[0048] Based on the smoke diffusion model, the time point when the fire smoke reaches the boundary of the adjacent fire compartment B is denoted as t2, the time point when it fills the preset proportion area in the fire compartment B is denoted as t3, and the time point when it fills the entire fire compartment B is denoted as t4.

[0049] Based on the loss model, the actual maximum power of the smoke exhaust fan in fire compartment B is calculated, and the time required for the smoke exhaust fan in fire compartment B to start and operate at its actual maximum power is recorded as T.

[0050] Calculate the difference between (t2-t0) and combine it with T to control the start-up time of the smoke exhaust fan in the fire compartment B to t1, so that when the fire smoke reaches the boundary of the fire compartment B, the smoke exhaust fan in the fire compartment B can work at its actual maximum power.

[0051] In the time interval [t1, t3], based on the first model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the first temperature. It is determined whether the first temperature exceeds the temperature threshold. If yes, a first-level warning is issued and the smoke exhaust fans in all adjacent fire compartments of fire compartment B are turned on. If no, in the time interval [t3, t4], based on the second model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the second temperature. It is determined whether the second temperature exceeds the temperature threshold. If yes, a second-level warning is issued and the smoke exhaust fans in a preset number of adjacent fire compartments of fire compartment B are turned on. If no, no action is taken. The first-level warning has a higher priority than the second-level warning.

[0052] A second aspect of this application provides a method for controlling a smoke exhaust fan, comprising:

[0053] A loss model is constructed and trained. Based on the trained loss model, the performance loss of all smoke exhaust fans in each fire compartment is calculated to obtain the loss calculation results.

[0054] Obtain the first coordinate curve of the maximum power of the exhaust fan as a function of temperature. Based on the loss calculation results, clean the data in the first coordinate curve. Based on the first coordinate curve after cleaning, determine the temperature corresponding to the maximum power drop of the exhaust fan and record it as the temperature threshold.

[0055] In the preset simulation software, a first model is constructed to measure the temperature change of the exhaust fan from startup to operation at maximum power in a smoke-free state; and a second model is constructed to measure the temperature change of the exhaust fan when it operates at maximum power in a smoke-containing state; the maximum power of the exhaust fan in the first model and the second model are the actual maximum power calculated by the loss model.

[0056] A smoke recognition model was constructed and trained. Based on the trained smoke recognition model, early smoke characteristics in each fire compartment were identified.

[0057] A smoke diffusion model was constructed and trained. Based on the trained smoke diffusion model, the diffusion of smoke in each fire compartment was simulated.

[0058] When a fire occurs in any fire compartment, the smoke exhaust fan in the corresponding fire compartment is turned on, and the data from the loss model, the temperature threshold, the first model, the second model and the smoke diffusion model are collected, analyzed and processed, and the smoke exhaust fans in the remaining fire compartments are precisely controlled.

[0059] Compared with the prior art, this application provides a smoke exhaust fan control system and method, including: a loss calculation unit, a temperature curve acquisition unit, a temperature change construction unit, a smoke monitoring unit, and a central control unit;

[0060] Since the performance of the exhaust fan will decrease due to daily use and wear and tear, this application uses a loss calculation unit to predict and calculate the performance of the exhaust fan after wear and tear, and uses it as the basic data of the exhaust fan (such as maximum power, maximum air volume, etc.) for subsequent modules.

[0061] The temperature threshold of the exhaust fan is obtained through the temperature curve acquisition unit. The basis for this is that the performance of the exhaust fan is affected as the temperature rises. When the temperature threshold is exceeded, the performance of the exhaust fan drops significantly. Therefore, this application obtains the first coordinate curve and uses the loss calculation results in the loss calculation unit to clean the data in the first coordinate curve. Then, the temperature threshold of the exhaust fan is obtained from the coordinate curve after data cleaning, with the maximum power drop as the indicator.

[0062] To achieve precise control of the smoke exhaust fan, this application also constructs a temperature change model in the preset simulation software. This model shows the temperature change of the smoke exhaust fan from startup to maximum power operation in a smoke-free state, and a second model shows the temperature change of the smoke exhaust fan at maximum power operation in a smoke-smoked state. It should be noted that the presence / absence of smoke is determined by the smoke encroaching on a preset proportion of the fire compartment. The reason for constructing the first and second models using simulation software is that if the smoke has not encroached on the preset proportion of the fire compartment, but the operating temperature of the smoke exhaust fan has already reached the temperature threshold (obtained through the first model), the warning level and handling measures are clearly different compared to when the smoke has encroached on the preset proportion of the fire compartment and the operating temperature of the smoke exhaust fan has reached the temperature threshold (obtained through the second model).

[0063] Since existing technologies only use smoke concentration to trigger fire alarms and activate subsequent smoke exhaust fans, they cannot immediately implement subsequent actions by detecting early smoke characteristics. Therefore, this application constructs and trains a smoke recognition model using a smoke monitoring unit. Based on the trained smoke recognition model, early smoke characteristics in each fire compartment are identified to detect fires and take subsequent measures in the shortest possible time.

[0064] This application simulates smoke diffusion by constructing a smoke diffusion unit, thereby obtaining the speed of smoke diffusion and the time required for smoke to invade other fire compartments at a predetermined proportion.

[0065] Finally, the central control unit integrates data from the loss calculation unit, temperature curve acquisition unit, temperature change construction unit, smoke monitoring unit, and smoke diffusion unit to achieve precise control of the smoke exhaust fan and improve the smoke exhaust efficiency of the entire fire smoke exhaust system.

[0066] Specifically, the performance of the smoke exhaust fan is predicted by the loss calculation unit, the operating threshold of the smoke exhaust fan is obtained by the temperature curve acquisition unit, the early smoke characteristics are identified by the smoke monitoring unit, the smoke diffusion rate and the time required for smoke to invade the preset proportion of areas in different fire compartments are obtained by the smoke diffusion unit, and then the first model and the second model are constructed by the temperature change construction unit to determine whether the smoke exhaust fan reaches the temperature threshold under different conditions, so as to carry out different levels of early warning and implement corresponding measures, so as to minimize the decrease in the smoke exhaust efficiency of the entire fire smoke exhaust system due to the performance degradation of the smoke exhaust fan caused by temperature.

[0067] Its beneficial effects are as follows: This application constructs a loss calculation unit, a temperature curve acquisition unit, a temperature change construction unit, a smoke monitoring unit, a smoke diffusion unit, and a central control unit. The central control unit integrates the data from the loss calculation unit, the temperature curve acquisition unit, the temperature change construction unit, the smoke monitoring unit, and the smoke diffusion unit. Specifically, the loss calculation unit predicts the working performance of the smoke exhaust fan, the temperature curve acquisition unit obtains the working threshold of the smoke exhaust fan, the smoke monitoring unit identifies early smoke characteristics, the smoke diffusion unit obtains the speed of smoke diffusion and the time required to invade preset proportions of different fire compartments, and the temperature change construction unit constructs a first model and a second model to determine whether the smoke exhaust fan reaches the temperature threshold under different conditions. This allows for different levels of early warning and the execution of corresponding measures, minimizing the decrease in smoke exhaust efficiency of the entire fire smoke exhaust system caused by the performance degradation of the smoke exhaust fan due to temperature. In other words, by achieving precise control of the smoke exhaust fan, the smoke exhaust efficiency of the entire fire smoke exhaust system is improved. Attached Figure Description

[0068] To more clearly illustrate the technical solutions in the embodiments of this application 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 embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0069] Figure 1 This is a schematic diagram of a smoke exhaust fan control system provided in an embodiment of this application;

[0070] Figure 2 This is a schematic flowchart of a smoke exhaust fan control method provided in an embodiment of this application. Detailed Implementation

[0071] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0072] In this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0073] Example 1

[0074] As can be seen from the background technology above, in large buildings, there are multiple fire compartments, and each fire compartment contains a smoke exhaust fan. When a fire occurs in a fire compartment, its corresponding smoke exhaust fan and the smoke exhaust fans of other fire compartments will start to work. However, the smoke exhaust fan will generate a lot of heat when it is working, and since the smoke also carries a lot of heat, the working environment temperature of the smoke exhaust fan will rise, thereby reducing the optimal performance of the smoke exhaust fan, resulting in a decrease in smoke exhaust volume and smoke exhaust efficiency, and ultimately affecting the smoke exhaust efficiency of the entire fire smoke exhaust system.

[0075] In existing technologies, when a certain concentration of smoke is detected by a smoke detector, the smoke exhaust fan in the corresponding fire compartment is turned on, and the smoke exhaust fans in other fire compartments are turned on simultaneously. However, this control method ignores the impact of temperature on the performance of the smoke exhaust fan, because the smoke exhaust fan generates a lot of heat when it is working, and the smoke it produces also carries a lot of heat. Therefore, how to reduce the impact of temperature on the performance of the smoke exhaust fan by accurately controlling the smoke exhaust fan, thereby improving the smoke exhaust efficiency of the entire fire smoke exhaust system.

[0076] Therefore, this application provides a smoke exhaust fan control system, such as... Figure 1 As shown, it includes:

[0077] Loss Calculation Unit: Constructs and trains a loss model, and based on the trained loss model, calculates the performance loss of all smoke exhaust fans in each fire compartment to obtain the loss calculation results;

[0078] Temperature curve acquisition unit: acquires the first coordinate curve of the maximum power of the exhaust fan as a function of temperature, cleans the data in the first coordinate curve based on the loss calculation results, and determines the temperature corresponding to the maximum power drop of the exhaust fan based on the cleaned first coordinate curve, and records it as the temperature threshold.

[0079] Temperature change construction unit: In the preset simulation software, a first model is constructed to model the temperature change of the exhaust fan from startup to operation at maximum power in a smoke-free state; and a second model is constructed to model the temperature change of the exhaust fan when it operates at maximum power in a smoke-containing state; the maximum power of the exhaust fan in the first model and the second model are the actual maximum power calculated by the loss model;

[0080] Smoke monitoring unit: Constructs and trains a smoke recognition model, and identifies early smoke characteristics in each fire compartment based on the trained smoke recognition model;

[0081] Smoke diffusion unit: Construct and train a smoke diffusion model, and simulate the diffusion of smoke in each fire compartment based on the trained smoke diffusion model;

[0082] Central control unit: When a fire occurs in any fire compartment, the smoke exhaust fan in the corresponding fire compartment is activated, and the data from the loss calculation unit, the temperature curve acquisition unit, the temperature change construction unit and the smoke diffusion unit are collected, analyzed and processed, and the smoke exhaust fans in the remaining fire compartments are precisely controlled.

[0083] Due to the influence of service life and working environment, the operating parameters of the exhaust fan will decrease to a certain extent compared with the calibrated parameters. Therefore, this embodiment constructs and trains a loss model. The construction logic of the loss model is to construct a set of loss functions for air volume, air pressure, maximum power, efficiency and energy consumption. The loss ratio of air volume, air pressure, maximum power, efficiency and energy consumption is calculated through the loss function set. Then, the above loss ratios are summed and their mean is calculated to obtain the loss coefficient of the exhaust fan. Then, the loss coefficient of the exhaust fan at multiple time scales is collected, and the coordinate curve of the loss system changing with time is constructed and fitted. The fitted coordinate curve data is used as the training set of the loss model to train the model. Then, the trained loss model is used to predict the performance loss of the exhaust fan.

[0084] Specifically, the performance loss of the exhaust fan is calculated based on the trained loss model. This involves inputting the usage time of the exhaust fan into the loss model, which then uses the fitted coordinate curve of the exhaust fan to predict the loss coefficient. The loss coefficient is then used to calculate the degree of loss of the exhaust fan, meaning that the calibration parameters and the loss coefficient are used as the actual parameters of the exhaust fan.

[0085] Before leaving the factory, exhaust fans undergo various performance tests, including air volume, air pressure, efficiency, and high-temperature testing. The high-temperature test measures the performance changes of the exhaust fan under high-temperature conditions. The first coordinate curve of the exhaust fan as a function of temperature can be obtained by compiling the factory test data. Alternatively, the performance parameters of the exhaust fan tested under normal conditions can be theoretically derived and converted to high-temperature conditions (see the journal "The Influence of Temperature on the Performance Curves and Characteristic Curves of Exhaust Fans and Ducts" for details). Then, based on the loss calculation results, that is, by calculating the product of the maximum rated power of the exhaust fan and the loss coefficient, the actual maximum power of the exhaust fan is obtained. The data in the first coordinate curve that exceed the actual maximum power are cleaned and processed. Then, the remaining first coordinate curve is divided into equal intervals, and the decrease in the actual maximum power in each interval (i.e., the slope of the curve) is calculated. That is, by using the mathematical concept of differentiation, when there are enough intervals, the data point with the largest curve slope can be obtained. In this embodiment, the data point with the temperature value at the left end of the interval is taken as the data point with the largest curve slope, that is, the temperature threshold point.

[0086] It should be noted that the reason for setting the concept of a temperature threshold in this embodiment is that after analyzing a large amount of historical data of the smoke exhaust fan, it was found that the performance of different smoke exhaust fans would drop significantly when the corresponding temperature threshold was exceeded. Therefore, how to control the smoke exhaust fan to prevent it from reaching the temperature threshold or reduce the working time of the smoke exhaust fan in an environment exceeding the temperature threshold, thereby improving the smoke exhaust efficiency of the smoke exhaust fan, is an urgent problem to be solved.

[0087] In the preset simulation software, a first model and a second model under two states are constructed. Modern simulation software can fully simulate the heat generation process and its thermodynamic behavior such as heat conduction, convection and radiation when the equipment is working (as well as multi-physics coupling, such as smoke scene, heat dissipation scene, etc.). The preset simulation software can be ANSYS Fluent / Mechanical, COMSOL Multiphysics and OpenFOAM, etc., and this embodiment does not make specific limitations.

[0088] The system includes a first model of temperature change of the smoke exhaust fan from startup to maximum power operation in a smoke-free state, and a second model of temperature change of the smoke exhaust fan operating at maximum power in a smoke-stricken state. It should be noted that the presence / absence of smoke is determined by a preset percentage of smoke coverage within the fire compartment. For example, if the preset percentage is set to 30%, the smoke exhaust fan is considered to be operating in a smoke-free state when smoke spreads or covers less than 30% of the fire compartment; and it is considered to be operating in a smoke-free state when smoke spreads or covers more than 30% of the fire compartment. The smoke exhaust fan operates under smoke conditions. The reason for constructing the first and second models using simulation software is that if the smoke exhaust fan's operating temperature has reached the temperature threshold (obtained through the first model) before the smoke has occupied 30% of the fire compartment, the warning level and handling measures will obviously differ from those if the smoke exhaust fan's operating temperature reaches the temperature threshold when the smoke has occupied 30% of the fire compartment (obtained through the second model). This embodiment uses the first and second models to determine the warning level and adopt handling measures, thereby achieving precise control of the smoke exhaust fan.

[0089] The process involves constructing and training a smoke recognition model to identify early smoke features. The model's construction and training employ deep learning, specifically: collecting surveillance video data from multiple time scales (e.g., one or two years) as a sample set; then augmenting the sample set by cropping, flipping, scaling, and translating the data to increase its size; manually labeling the augmented sample set with attributes including, but not limited to, smoke location and size, confidence scores for whether a smoke area triggers an alarm, smoke concentration, and fire severity; and finally, using a Region Candidate Network (RPN) to generate candidate regions containing smoke, followed by classification and regression training using a Convolutional Neural Network (CNN) to output the smoke category and location coordinates.

[0090] This involves constructing a smoke diffusion model to simulate indoor smoke diffusion. During the simulation, to calculate the values ​​of physical quantities such as smoke velocity, pressure, density, and temperature at any time step, initial parameters need to be set for the smoke model solution. These parameters include the BIM data of the fire compartment, the location of the smoke source, and the initial values ​​of the smoke source's physical quantities (velocity, density, and temperature, etc.). To enhance the detail of the smoke simulation, wind force, diffusion force, and air resistance are treated as external forces, and separate models for wind force, diffusion force, and air resistance are established and solved. The construction and solution of these wind force, diffusion force, and air resistance models are discussed in detail. The solution is already a relatively mature technology for those skilled in the art, and will not be elaborated on in this embodiment. Then, based on the set initial parameters and the external force obtained by the solution, the smoke diffusion model is solved. That is, at each time step, the velocity field, temperature field and density field of the smoke are updated. Then, based on the size of the fire compartment and the boundary of the obstacle, the changes in velocity, temperature and density of the smoke after encountering the wall and obstacle are calculated, thereby correcting and constraining the movement of the smoke. Based on the smoke density field, velocity field and temperature field calculated by the smoke diffusion model, the diffusion of the smoke is simulated and plotted in real time.

[0091] The central control unit analyzes and processes the data from the loss calculation unit, temperature curve acquisition unit, temperature change construction unit, and smoke diffusion unit. The smoke monitoring unit identifies the presence of fire smoke in fire compartment A and activates the smoke exhaust fan in fire compartment A. At the same time, the loss calculation unit calculates the loss coefficient of the smoke exhaust fans in all fire compartments and calculates the actual maximum efficiency of the smoke exhaust fans in all fire compartments based on the loss coefficient. The temperature curve acquisition unit also obtains the temperature threshold of the smoke exhaust fans in all fire compartments.

[0092] Based on the smoke diffusion model, the time point when the fire smoke reaches the boundary of the adjacent fire compartment B is t2, the time point when it fills the preset proportion area (30%) in the fire compartment B is t3, and the time point when it fills the entire fire compartment B is t4.

[0093] Based on the loss model, the actual maximum power of the smoke exhaust fan in fire compartment B is calculated, and the time required for the smoke exhaust fan in fire compartment B to start and operate at its actual maximum power is recorded as T.

[0094] Calculate the difference between t2 and t0, and combine it with T to control the start-up time of the smoke exhaust fan in fire compartment B to t1, i.e., t2-t1≥T, preferably t2-t1=T. This means that when the fire smoke reaches the boundary of fire compartment B, the smoke exhaust fan in fire compartment B can work at maximum power. This can ensure smoke exhaust efficiency and reduce the large amount of heat generated by starting the smoke exhaust fan in advance, which would affect the working performance of the smoke exhaust fan.

[0095] In the time interval [t1, t3], based on the first model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the first temperature. It is then determined whether the first temperature exceeds the temperature threshold. If so, a level one warning is issued and the smoke exhaust fans in all adjacent fire compartments of fire compartment B are activated. This indicates that when fire smoke only invades a small area of ​​fire compartment B, the performance of the smoke exhaust fan is significantly affected. If the smoke exhaust fans in all adjacent areas are not activated in advance, the smoke exhaust efficiency will decrease as the fire smoke invades more areas, leading to greater safety hazards. If not, in the time interval [t3, t4], based on the second model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the first temperature. It is then determined whether the first temperature exceeds the temperature threshold. If so, a level one warning is issued and the smoke exhaust fans in all adjacent fire compartments of fire compartment B are activated. The maximum temperature that the smoke exhaust fan in fire compartment B can reach is recorded as the second temperature. It is determined whether the second temperature exceeds the temperature threshold. If it does, a level two warning is issued and the smoke exhaust fans in a preset number (usually one or two) of the adjacent fire compartments of fire compartment B are turned on. This indicates that the performance of the smoke exhaust fans is only significantly affected when fire smoke invades a large area of ​​fire compartment B. Therefore, only one or two smoke exhaust fans in the fire compartments need to be turned on in advance to ensure the smoke exhaust efficiency of the entire smoke exhaust fire protection system and to minimize the time when the performance of the smoke exhaust fans declines due to operating temperature. If not, no action is taken. The level one warning has a higher priority than the level two warning.

[0096] In some embodiments, constructing and training the loss model includes:

[0097] Construct a set of loss functions for the air volume, air pressure, maximum power, efficiency, and energy consumption of the exhaust fan, specifically:

[0098]

[0099] This represents the percentage of air volume loss. This is the weighting coefficient for air volume. This refers to the actual air volume. For the target air volume, This represents the proportion of wind pressure loss. This is the weighting coefficient for wind pressure. This is the actual wind pressure. For target wind pressure, This represents the percentage of maximum power loss. The weighting factor for maximum power. This is the actual maximum power. For the target maximum power, This represents the percentage loss in conversion efficiency. The weighting coefficient for conversion efficiency. For actual conversion efficiency, To achieve the target conversion efficiency, This represents the proportion of energy loss. This is a weighting factor for energy consumption. Actual energy consumption Target energy consumption;

[0100] Based on the aforementioned set of loss functions, the loss coefficients of the exhaust fan are constructed as follows:

[0101]

[0102] The loss coefficients of the exhaust fan at multiple time scales are obtained, and the results are obtained.

[0103] Based on the obtained results, a coordinate curve showing the change of the loss coefficient with the usage time of the exhaust fan is constructed with the usage time as the horizontal axis and the loss coefficient of the exhaust fan as the vertical axis, and then fitted.

[0104] In some embodiments, performance loss calculations are performed on all smoke exhaust fans in each fire compartment, and the loss calculation results include:

[0105] The usage time of any smoke exhaust fan is recorded as the first usage time, and the fitted coordinate curve is recorded as the second coordinate curve.

[0106] Based on the fitted second coordinate curve, the loss coefficient at the first usage time node is predicted and denoted as the first loss coefficient.

[0107] Based on the first loss coefficient, the loss calculation results of the corresponding smoke exhaust fan are obtained.

[0108] In some embodiments, cleaning the data in the first coordinate curve based on the loss calculation results includes:

[0109] Based on the loss calculation results, the actual maximum power of the exhaust fan is obtained;

[0110] In the first coordinate curve, the data point corresponding to the actual maximum power is matched, and the curve data located before the data point is cleaned.

[0111] In some embodiments, determining the temperature threshold based on the first coordinate curve after cleaning includes:

[0112] In the first coordinate curve, taking the data point corresponding to the actual maximum power as the starting point, the first coordinate curve located after the data point is divided into several equal intervals, and the rate of change of the first coordinate curve in each interval is calculated to obtain the rate of change calculation result.

[0113] Based on the calculation results of the rate of change, the interval corresponding to the largest rate of change is selected and denoted as the first interval;

[0114] The temperature value at the left port of the first interval is recorded as the temperature threshold.

[0115] In some embodiments, constructing and training a smoke recognition model includes:

[0116] Collect surveillance video data at multiple time scales as a sample set and store it in the database;

[0117] Data augmentation is performed on the sample set in the database to expand the data volume of the sample set;

[0118] Manually label the data-enhanced sample set;

[0119] Based on a manually labeled sample set, a region candidate network is used to generate candidate regions containing smoke.

[0120] A convolutional neural network is used to classify and regress the candidate region, and output the category and location coordinates of the smoke.

[0121] In some embodiments, constructing and training a smoke diffusion model includes:

[0122] Construct a smoke diffusion model and set initial parameters for the smoke diffusion model, including BIM data of the fire compartment, location of the smoke source, and physical quantities of the smoke source;

[0123] Establish wind force model, diffusion force model and air resistance model and solve them separately. Combine the solution results to obtain external force data.

[0124] Based on the initial parameters and the merged external force data, the smoke diffusion model is solved, and the velocity field, temperature field and density field of the smoke are updated.

[0125] Based on the size of the fire compartment and the boundaries of obstacles, the movement of smoke is constrained, thereby correcting the velocity field, temperature field and density field of the smoke;

[0126] Based on the modified velocity field, temperature field, and density field of the smoke in the smoke diffusion model, the diffusion of smoke is simulated and plotted in real time.

[0127] In some embodiments, when a fire occurs in any fire compartment, activating the smoke exhaust fan in the corresponding fire compartment includes:

[0128] Based on the smoke monitoring unit, the type of smoke and the coordinates of its location in any fire compartment can be obtained;

[0129] If the smoke is determined to be fire smoke, the smoke exhaust fan of the corresponding fire compartment is turned on, and the location coordinates of the fire smoke are uploaded to the preset terminal.

[0130] In some embodiments, precise control of smoke exhaust fans in the remaining fire compartments includes:

[0131] When there is fire smoke in fire compartment A, the time point at which the fire smoke occurs is recorded as t0.

[0132] Based on the smoke diffusion model, the time point when the fire smoke reaches the boundary of the adjacent fire compartment B is denoted as t2, the time point when it fills the preset proportion area in the fire compartment B is denoted as t3, and the time point when it fills the entire fire compartment B is denoted as t4.

[0133] Based on the loss model, the actual maximum power of the smoke exhaust fan in fire compartment B is calculated, and the time required for the smoke exhaust fan in fire compartment B to start and operate at its actual maximum power is recorded as T.

[0134] Calculate the difference between (t2-t0) and combine it with T to control the start-up time of the smoke exhaust fan in the fire compartment B to t1, so that when the fire smoke reaches the boundary of the fire compartment B, the smoke exhaust fan in the fire compartment B can work at its actual maximum power.

[0135] In the time interval [t1, t3], based on the first model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the first temperature. It is determined whether the first temperature exceeds the temperature threshold. If yes, a first-level warning is issued and the smoke exhaust fans in all adjacent fire compartments of fire compartment B are turned on. If no, in the time interval [t3, t4], based on the second model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the second temperature. It is determined whether the second temperature exceeds the temperature threshold. If yes, a second-level warning is issued and the smoke exhaust fans in a preset number of adjacent fire compartments of fire compartment B are turned on. If no, no action is taken. The first-level warning has a higher priority than the second-level warning.

[0136] Example 2

[0137] Based on the smoke exhaust fan control system provided in Embodiment 1 of this application, correspondingly, Embodiment 2 of this application also provides a smoke exhaust fan control method, such as... Figure 2 As shown, it includes:

[0138] A loss model is constructed and trained. Based on the trained loss model, the performance loss of all smoke exhaust fans in each fire compartment is calculated to obtain the loss calculation results.

[0139] Obtain the first coordinate curve of the maximum power of the exhaust fan as a function of temperature. Based on the loss calculation results, clean the data in the first coordinate curve. Based on the first coordinate curve after cleaning, determine the temperature corresponding to the maximum power drop of the exhaust fan and record it as the temperature threshold.

[0140] In the preset simulation software, a first model is constructed to measure the temperature change of the exhaust fan from startup to operation at maximum power in a smoke-free state; and a second model is constructed to measure the temperature change of the exhaust fan when it operates at maximum power in a smoke-containing state; the maximum power of the exhaust fan in the first model and the second model are the actual maximum power calculated by the loss model.

[0141] A smoke recognition model was constructed and trained. Based on the trained smoke recognition model, early smoke characteristics in each fire compartment were identified.

[0142] A smoke diffusion model was constructed and trained. Based on the trained smoke diffusion model, the diffusion of smoke in each fire compartment was simulated.

[0143] When a fire occurs in any fire compartment, the smoke exhaust fan in the corresponding fire compartment is turned on, and the data from the loss model, the temperature threshold, the first model, the second model and the smoke diffusion model are collected, analyzed and processed, and the smoke exhaust fans in the remaining fire compartments are precisely controlled.

[0144] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computing software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0145] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0146] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A smoke exhaust fan control system, characterized in that, include: Loss Calculation Unit: Constructs and trains a loss model, and based on the trained loss model, calculates the performance loss of all smoke exhaust fans in each fire compartment to obtain the loss calculation results; Temperature curve acquisition unit: acquires the first coordinate curve of the maximum power of the exhaust fan as a function of temperature, cleans the data in the first coordinate curve based on the loss calculation results, and determines the temperature corresponding to the maximum power drop of the exhaust fan based on the cleaned first coordinate curve, and records it as the temperature threshold. Temperature change construction unit: In the preset simulation software, a first model is constructed to show the temperature change of the exhaust fan from startup to operation at maximum power in a smoke-free state. And a second model is constructed to illustrate the temperature change of the exhaust fan when it operates at maximum power in the presence of smoke; the maximum power of the exhaust fan in both the first and second models is the actual maximum power calculated through the loss model; Smoke monitoring unit: Constructs and trains a smoke recognition model, and identifies early smoke characteristics in each fire compartment based on the trained smoke recognition model; Smoke diffusion unit: Construct and train a smoke diffusion model, and simulate the diffusion of smoke in each fire compartment based on the trained smoke diffusion model; Central Control Unit: When a fire occurs in any fire compartment, the smoke exhaust fan in the corresponding fire compartment is activated, and data from the loss calculation unit, temperature curve acquisition unit, temperature change construction unit, and smoke diffusion unit are collected, analyzed, and processed to precisely control the smoke exhaust fans in the remaining fire compartments, specifically: When there is fire smoke in fire compartment A, the time point at which the fire smoke occurs is recorded as t0. Based on the smoke diffusion model, the time point when the fire smoke reaches the boundary of the adjacent fire compartment B is denoted as t2, the time point when it fills the preset proportion area in the fire compartment B is denoted as t3, and the time point when it fills the entire fire compartment B is denoted as t4. Based on the loss model, the actual maximum power of the smoke exhaust fan in fire compartment B is calculated, and the time required for the smoke exhaust fan in fire compartment B to start and operate at its actual maximum power is recorded as T. Calculate the difference between (t2-t0) and combine it with T to control the start-up time of the smoke exhaust fan in the fire compartment B to t1, so that when the fire smoke reaches the boundary of the fire compartment B, the smoke exhaust fan in the fire compartment B can work at its actual maximum power. In the time interval [t1, t3], based on the first model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the first temperature. It is determined whether the first temperature exceeds the temperature threshold. If yes, a first-level warning is issued and the smoke exhaust fans in all adjacent fire compartments of fire compartment B are turned on. If no, in the time interval [t3, t4], based on the second model, the maximum temperature that the smoke exhaust fan in fire compartment B can reach is output, denoted as the second temperature. It is determined whether the second temperature exceeds the temperature threshold. If yes, a second-level warning is issued and the smoke exhaust fans in a preset number of adjacent fire compartments of fire compartment B are turned on. If no, no action is taken. The first-level warning has a higher priority than the second-level warning.

2. The smoke exhaust fan control system according to claim 1, characterized in that, Building and training a loss model includes: Construct a set of loss functions for the air volume, air pressure, maximum power, efficiency, and energy consumption of the exhaust fan, specifically: This represents the percentage of air volume loss. This is the weighting coefficient for air volume. This refers to the actual air volume. For the target air volume, This represents the proportion of wind pressure loss. This is the weighting coefficient for wind pressure. This is the actual wind pressure. For target wind pressure, This represents the percentage of maximum power loss. The weighting factor for maximum power. This is the actual maximum power. For the target maximum power, This represents the percentage loss in conversion efficiency. The weighting coefficient for conversion efficiency. For actual conversion efficiency, To achieve the target conversion efficiency, This represents the proportion of energy loss. This is a weighting factor for energy consumption. Actual energy consumption Target energy consumption; Based on the aforementioned set of loss functions, the loss coefficients of the exhaust fan are constructed as follows: The loss coefficients of the exhaust fan at multiple time scales are obtained, and the results are obtained. Based on the obtained results, a coordinate curve showing the change of the loss coefficient with the usage time of the exhaust fan is constructed with the usage time as the horizontal axis and the loss coefficient of the exhaust fan as the vertical axis, and then fitted.

3. The smoke exhaust fan control system according to claim 2, characterized in that, Performance loss calculations were performed on all smoke exhaust fans in each fire compartment, and the results included: The usage time of any smoke exhaust fan is recorded as the first usage time, and the fitted coordinate curve is recorded as the second coordinate curve. Based on the fitted second coordinate curve, the loss coefficient at the first usage time node is predicted and denoted as the first loss coefficient. Based on the first loss coefficient, the loss calculation results of the corresponding smoke exhaust fan are obtained.

4. The smoke exhaust fan control system according to claim 1, characterized in that, Based on the loss calculation results, the data cleaning process in the first coordinate curve includes: Based on the loss calculation results, the actual maximum power of the exhaust fan is obtained; In the first coordinate curve, the data point corresponding to the actual maximum power is matched, and the curve data located before the data point is cleaned.

5. A smoke exhaust fan control system according to claim 4, characterized in that, Based on the first coordinate curve after cleaning, the temperature threshold is determined as follows: In the first coordinate curve, taking the data point corresponding to the actual maximum power as the starting point, the first coordinate curve located after the data point is divided into several equal intervals, and the rate of change of the first coordinate curve in each interval is calculated to obtain the rate of change calculation result. Based on the calculation results of the rate of change, the interval corresponding to the largest rate of change is selected and denoted as the first interval; The temperature value at the left port of the first interval is recorded as the temperature threshold.

6. The smoke exhaust fan control system according to claim 1, characterized in that, Building and training a smoke recognition model includes: Collect surveillance video data at multiple time scales as a sample set and store it in the database; Data augmentation is performed on the sample set in the database to expand the data volume of the sample set; Manually label the data-enhanced sample set; Based on a manually labeled sample set, a region candidate network is used to generate candidate regions containing smoke. A convolutional neural network is used to classify and regress the candidate region, and output the category and location coordinates of the smoke.

7. The smoke exhaust fan control system according to claim 1, characterized in that, Building and training a smoke diffusion model includes: Construct a smoke diffusion model and set initial parameters for the smoke diffusion model, including BIM data of the fire compartment, location of the smoke source, and physical quantities of the smoke source; Establish wind force model, diffusion force model and air resistance model and solve them separately. Combine the solution results to obtain external force data. Based on the initial parameters and the merged external force data, the smoke diffusion model is solved, and the velocity field, temperature field and density field of the smoke are updated. Based on the size of the fire compartment and the boundaries of obstacles, the movement of smoke is constrained, thereby correcting the velocity field, temperature field and density field of the smoke; Based on the modified velocity field, temperature field, and density field of the smoke in the smoke diffusion model, the diffusion of smoke is simulated and plotted in real time.

8. The smoke exhaust fan control system according to claim 1, characterized in that, When a fire occurs in any fire compartment, activating the smoke exhaust fan in that fire compartment includes: Based on the smoke monitoring unit, the type of smoke and the coordinates of its location in any fire compartment can be obtained; If the smoke is determined to be fire smoke, the smoke exhaust fan of the corresponding fire compartment is turned on, and the location coordinates of the fire smoke are uploaded to the preset terminal.

9. A method for controlling a smoke exhaust fan, implemented by a smoke exhaust fan control system as described in claim 1, characterized in that, include: A loss model is constructed and trained. Based on the trained loss model, the performance loss of all smoke exhaust fans in each fire compartment is calculated to obtain the loss calculation results. Obtain the first coordinate curve of the maximum power of the exhaust fan as a function of temperature. Based on the loss calculation results, clean the data in the first coordinate curve. Based on the first coordinate curve after cleaning, determine the temperature corresponding to the maximum power drop of the exhaust fan and record it as the temperature threshold. In the preset simulation software, a first model is constructed to simulate the temperature change of the exhaust fan from startup to operation at maximum power in a smoke-free state. And a second model is constructed to illustrate the temperature change of the exhaust fan when it operates at maximum power in the presence of smoke; the maximum power of the exhaust fan in both the first and second models is the actual maximum power calculated through the loss model; A smoke recognition model was constructed and trained. Based on the trained smoke recognition model, early smoke characteristics in each fire compartment were identified. A smoke diffusion model was constructed and trained. Based on the trained smoke diffusion model, the diffusion of smoke in each fire compartment was simulated. When a fire occurs in any fire compartment, the smoke exhaust fan in the corresponding fire compartment is turned on, and the data from the loss model, the temperature threshold, the first model, the second model and the smoke diffusion model are collected, analyzed and processed, and the smoke exhaust fans in the remaining fire compartments are precisely controlled.

Images