Pipe gallery fire emergency ventilation and smoke exhaust system

Abstract

This invention relates to the field of fire emergency management technology, specifically disclosing a ventilation and smoke extraction system for pipe gallery fires. It simultaneously collects concentration, temperature, visibility, micro-pressure difference in gaps, and bidirectional wind speed at both sides of the fire compartment boundary. Concentration bias and temperature bias are calculated, and weighted fusion of multiple feature values ​​generates a boundary fire spread index. When the index exceeds a threshold, an adaptive control loop is triggered: if both biases are below the symmetry threshold, a dual-zone collaborative mode is activated, dynamically adjusting the frequency of the fans on both sides with zero wind speed as the target; if the concentration bias exceeds the unidirectional threshold, the fire compartment is locked and a unilateral defense mode is activated, establishing a wind pressure barrier based on the pressure difference deviation. If the time derivative of the micro-pressure difference exceeds the oscillation threshold during the loop, the frequency of all fans is frozen until damping ends, thereby preventing airflow resonance and achieving efficient directional isolation and extraction of smoke from pipe gallery fires.

Description

Technical Field

[0001] This invention belongs to the field of fire emergency management technology and relates to a fire emergency ventilation and smoke exhaust system for pipe corridors. Background Technology

[0002] As a crucial component of urban underground space, utility tunnels serve as integrated installations for various municipal pipelines, including power, communications, water supply and drainage, and gas. Their ventilation and smoke extraction systems not only ensure the safe operation of equipment within the tunnels but also directly impact the overall effectiveness of smoke control, personnel evacuation, and fire suppression during fires. For long-distance utility tunnels, a zoned ventilation structure is commonly used. Each fire compartment typically has its own independent intake and exhaust fans to achieve segmented air supply and exhaust, zoned air exchange, and localized environmental regulation during normal operation, thereby maintaining stable temperature, humidity, pollutant concentration, and airflow organization within the tunnel.

[0003] Furthermore, when a fire occurs at the boundary between two fire compartments, because the fire source is close to the fire door or compartment boundary, smoke often simultaneously intrudes into both adjacent areas, forming a complex smoke field across compartments. In this situation, simply relying on the smoke extraction equipment in the fire compartment can extract some local smoke, but because the adjacent compartments are still under normal ventilation, the boundary will continue to be affected by pressure differences and airflow recirculation, causing smoke to repeatedly migrate, stagnate, and even oscillate and diffuse between the two compartments. If the control strategy cannot accurately identify the degree of fire source deviation, the difference in smoke concentration on both sides of the boundary, and the airflow direction at the door gaps, a conflict can easily arise where smoke is extracted on one side while air is supplied on the other. This causes smoke that should be quickly removed to circulate repeatedly between compartments, reducing smoke extraction efficiency, prolonging the residence time of high-temperature toxic smoke in the pipe gallery, and increasing secondary risks near fire doors and adjacent compartments.

[0004] In boundary fire scenarios, the start-up and shutdown of fans, the switching of supply and exhaust air, and the status of fire doors often adopt a fixed sequence, which makes it difficult to dynamically adapt to the real-time changes in smoke concentration, temperature gradient, and door gap wind speed. This can easily cause smoke to oscillate back and forth between zones. Thirdly, the existing solution does not make full use of the reverse suction capacity of smoke exhaust fans in adjacent zones, nor does it refine the control of air volume matching under different boundary fire conditions. This results in problems such as rough control, delayed response, and obvious local backflow in the smoke exhaust process. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides an emergency ventilation and smoke exhaust system for pipe gallery fires to solve the above-mentioned technical problems.

[0006] To achieve the above and other objectives, the technical solution adopted by the present invention is as follows: This invention provides an emergency ventilation and smoke extraction system for pipe gallery fires. The system includes a first fire compartment, a second fire compartment, and a sensor array and air supply and exhaust equipment located at the boundary between the two compartments. The embedded controller performs the following smoke extraction and coordinated processing steps: Acquire the first concentration value, the second concentration value, the first temperature value, the second temperature value, the first visibility value, the second visibility value, the micro-pressure difference value and the bidirectional wind speed value at the junction gap on both sides; The difference between the first concentration value and the second concentration value is calculated and divided by the maximum value of the two to generate the concentration bias; the difference between the first temperature value and the second temperature value is calculated and divided by the maximum value of the two to generate the temperature bias. Calculate the sum of the first concentration value and the second concentration value, the sum of the first temperature value and the second temperature value, and the sum of the first visibility value and the second visibility value. Multiply each sum by its corresponding preset weight and sum them to generate the boundary fire spread index. An adaptive control loop is triggered when the boundary fire spread index exceeds a safety threshold. The adaptive control loop includes: If the absolute values ​​of both concentration bias and temperature bias are less than the symmetry threshold, the dual-zone collaborative mode is triggered, and the highest frequency command is output to control the operation of the first and second exhaust fans. The deviation between the bidirectional wind speed value and the zero value is calculated, and the operating frequency of the first and second exhaust fans is dynamically adjusted synchronously according to the deviation until the bidirectional wind speed value converges to the target wind speed zone. If the absolute value of the concentration bias is greater than the one-way bias threshold, the one-sided defense mode is triggered. The positive or negative polarity of the concentration bias is determined to lock the fire zone. The output command controls the main exhaust fan of the fire zone to run at the highest frequency. The pressure difference deviation between the micro pressure difference value and the target smoke control pressure difference value is calculated. Based on the pressure difference deviation, the operating frequency of the adjacent supply fan in the non-fire zone is adjusted to establish a wind pressure barrier. Within the adaptive control cycle, the time-varying derivative of the micro-pressure difference is calculated. If the time-varying derivative is greater than the airflow oscillation threshold, the damping lockout procedure is triggered, the calculation of the new adjustment frequency of each blower and exhaust fan is stopped, and the equipment is frozen at the current operating frequency until the damping countdown ends.

[0007] As described above, the emergency ventilation and smoke exhaust system for pipe gallery fires provided by the present invention has at least the following beneficial effects: The present invention uses the boundary fire spread index as the unified trigger condition for entering the control loop, and starts the dual-zone collaborative mode when both the absolute value of the concentration bias and the absolute value of the temperature bias are less than the symmetry threshold. In this mode, the exhaust fans on both sides are synchronously controlled to operate at full load, and at the same time, the supply fans on both sides are dynamically adjusted according to the deviation between the bidirectional wind speed value and zero, so as to form a more stable air flow convergence and exhaust path in the junction area; when the absolute value of the concentration bias is greater than the unidirectional bias threshold, it switches to the single-side defense mode, locks the fire area through the positive and negative polarities of the concentration bias, and takes the deviation between the differential pressure value and the target smoke control differential pressure value as the control basis to adjust the operating frequency of the supply fan in the non-fire area to establish a wind pressure barrier. The above control method no longer adopts the extensive smoke exhaust logic of fixed air volume and fixed time sequence, but incorporates the fire distribution, boundary pressure difference and air flow direction into the same control link for linkage judgment, enabling the system to automatically switch modes according to different evolution forms of the junction fire, significantly improving the accuracy, timeliness and stability of smoke control.

[0008] The present invention also introduces a monitoring mechanism for the time derivative of the differential pressure value within the adaptive control cycle. When it is detected that the time derivative is greater than the air flow oscillation threshold, the damping locking program is immediately triggered, and the new adjustment frequencies of the supply fans and exhaust fans are stopped and the equipment is frozen at the current operating frequency until the damping countdown ends and then the adjustment is resumed. This mechanism can suppress the system oscillation caused by frequent frequency modulation, wind direction reversal and junction air flow sawing, avoid the repeated migration and secondary diffusion of smoke between the two fire prevention zones, and transform the smoke exhaust process from the original passive following control into a stable closed-loop control with vibration suppression ability. Through this control scheme that links boundary fire recognition, mode switching, differential pressure barrier construction and damping locking, the present invention not only improves the smoke exhaust efficiency in the junction fire scenario, but also enhances the adaptability and reliability of the system in complex boundary conditions, thereby improving the overall safety guarantee ability of the pipe gallery fire emergency ventilation and smoke exhaust system. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0010] Figure 1 It is a schematic diagram of the system logic connection of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0011] The following description, in conjunction with the implementation of this invention, is merely an example and illustration of the concept of this invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in these claims, all of which should fall within the protection scope of this invention.

[0012] In traditional, existing emergency ventilation and smoke extraction systems for utility tunnel fires, fixed threshold values ​​and single-zone control rules are ill-suited to the rapid changes in smoke diffusion during boundary fires. When a fire occurs at the boundary of a fire compartment, smoke concentration, temperature, visibility, and airflow patterns at gaps often exhibit significant nonlinear fluctuations. If the system continues to use static threshold values ​​to determine the boundary fire situation, it cannot establish a dynamic correlation between various monitoring data and the actual fire evolution. This can easily lead to inaccurate identification of smoke deviation, delayed switching of smoke extraction modes, and mismatch between the supply and exhaust air directions and the on-site airflow conditions. This static control method reduces the accuracy of multi-source monitoring data fusion, causing invalid disturbances to be introduced into the characteristic quantities input to the controller, ultimately affecting the timeliness and relevance of the smoke extraction strategy.

[0013] For example, in a boundary fire scenario, if the first and second fire compartments experience smoke intrusion to varying degrees within the same sampling period, the first and second concentration values, the first and second temperature values, as well as the first and second visibility values, will change simultaneously. The micro-pressure difference and bidirectional wind speed at the boundary gap will also change accordingly. If the system still uses a fixed airflow and fixed timing for smoke extraction, it will be difficult to promptly distinguish whether the fire source is spreading symmetrically, and it will be difficult to determine whether the smoke is shifting along one side of the compartment or oscillating between the two compartments. This may lead to insufficient smoke extraction on one side and excessive air supply on the other side, causing smoke that should be quickly removed to repeatedly linger at the boundary, reducing smoke extraction efficiency and increasing the risk of cross-zone spread.

[0014] If the above problems are not addressed, misjudgment of smoke in boundary fire scenarios will lead to a misalignment between control strategies and the actual fire evolution stage, further exacerbating the circulation and migration of high-temperature toxic smoke between adjacent zones and prolonging the hazardous exposure time near fire doors. The rigidity of the feature screening mechanism will also hinder the system from capturing critical changes from symmetrical spread to unilateral bias, causing smoke exhaust equipment to maintain its original operating state when it should be switching over, delaying the optimal time to establish wind pressure barriers and stabilize airflow organization. At the same time, if the temporal variation of micro-pressure differences is not effectively constrained, frequent adjustments of supply and exhaust fans may induce airflow oscillations, causing the smoke boundary to swing back and forth, ultimately forming a negative feedback loop and affecting the iterative optimization of emergency ventilation and smoke exhaust strategies in boundary fires.

[0015] When faced with the aforementioned problems, traditional systems use fixed thresholds to filter cross-zone feature correlations, leading to the misclassification of key concentration, temperature, and airflow parameters as redundant fluctuations, failing to reflect the true spread trend of the fire at the boundary. To address this, this application attempts to dynamically couple concentration bias, temperature bias, and the boundary fire spread index, adjusting the internal criteria of the multi-dimensional data fusion model through real-time coefficient adjustments. Further analysis reveals that relying solely on single boundary data or static mapping relationships is insufficient to capture nonlinear fluctuations under boundary fires; therefore, it is necessary to introduce environmental data from both sides to collaboratively calculate the spread index and generate a dynamic control threshold based on its changing trends. By designing a linked structure of adaptive control loop and damping locking mechanism, the smoke exhaust mode adaptively switches according to the smoke intrusion state in the boundary area, thereby solving the problems of control lag, airflow oscillation, and boundary fire identification mismatch.

[0016] After introducing the basic concept of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0017] Example 1: Please refer to Figure 1 As shown, a fire emergency ventilation and smoke exhaust system for a utility tunnel includes a first fire compartment, a second fire compartment, and a sensor array and air supply and exhaust equipment located at the boundary between the two compartments. The embedded controller performs the following smoke exhaust coordinated processing steps: Acquire the first concentration value, second concentration value, first temperature value, second temperature value, first visibility value, and second visibility value on both sides of the boundary, as well as the micro-pressure difference value and bidirectional wind speed value at the boundary gap; including: Multimodal raw sensor signals from both sides and gaps at the interface are collected synchronously according to the preset sampling frequency, and the multimodal raw sensor signals are time-series aligned according to time points to construct a time-synchronized dataset. The first environmental electrical signal corresponding to the first fire compartment, the second environmental electrical signal corresponding to the second fire compartment, and the gap fluid electrical signal corresponding to the boundary gap are separated from the time synchronization dataset. Extract the numerical sets of the first environmental electrical signal, the second environmental electrical signal, and the gap fluid electrical signal within the sliding time window and calculate the average value to generate the first smoothed signal, the second smoothed signal, and the gap smoothed signal after filtering out high-frequency noise, respectively. The first smoothed signal is calibrated and mapped using physical quantities to obtain the first concentration value, the first temperature value, and the first visibility value. The second smoothed signal is calibrated and mapped using physical quantities to obtain the second concentration value, the second temperature value, and the second visibility value. The environmental pressure difference feature and airflow movement feature are extracted by decoupling the gap smoothing signal. The micro pressure difference value is calculated based on the environmental pressure difference feature, and the airflow movement feature is combined with the airflow movement direction factor to generate bidirectional wind speed value through vector calculation.

[0018] In this embodiment of the invention, from the time synchronization dataset, according to a preset hardware port mapping protocol, the first environmental electrical signal sequence corresponding to the first fire compartment (including the voltage response values ​​of temperature, concentration, and visibility sensors), the second environmental electrical signal sequence corresponding to the second fire compartment, and the gap fluid electrical signal sequence corresponding to the boundary gap (including the voltage response values ​​of the differential pressure transmitter and the hot wire anemometer) are separated respectively.

[0019] Subsequently, the first environmental electrical signal, the second environmental electrical signal, and the gap fluid electrical signal are subjected to noise reduction processing using a sliding time window. The numerical sets within the window are extracted and their arithmetic averages are calculated to generate a first smoothed signal, a second smoothed signal, and a gap smoothed signal with high-frequency noise filtered out. Further, the first smoothed signal and the second smoothed signal are subjected to physical quantity calibration mapping, using the built-in linear sensor response curve equation. The smoothed voltage signal is analyzed to extract a first concentration value, a first temperature value, a first visibility value, and a second concentration value, a second temperature value, and a second visibility value, all with clearly defined physical dimensions; among which... Input voltage, This is the zero-point bias voltage under no-environment excitation conditions. To calibrate the sensitivity coefficient, the above reference parameters were all obtained directly from the factory calibration and written into the controller's non-volatile memory.

[0020] For the extremely complex bidirectional flue gas intrusion flow at the gaps, the system includes differential pressure and voltage values. and heat dissipation wind speed voltage value The gap smoothing signal is decoupled based on its characteristics. First, the micro-pressure difference value is calculated based on the characteristics of the environmental pressure difference. ,in This is the pressure difference conversion factor, and its dimensions are... , The resting state reference voltage value is used to output a voltage with positive and negative polarities and dimensions. The micro-pressure difference value is calculated, where a positive value indicates that the pressure in the first zone is relatively high. Subsequently, airflow movement characteristics are combined with an airflow direction factor to perform vector calculations, generating bidirectional wind speed values ​​that include both the magnitude and precise direction of the flow velocity. To overcome the technical shortcomings of traditional King's Law in determining airflow direction and failing to consider the frictional interference effect of extremely narrow fire door gaps, a calculation formula is used: In this formula, This is the final output of the bidirectional wind speed value; The sign function is used to extract the direction of the airflow movement, which takes the micro-pressure difference polarity as the airflow movement direction factor. The result returns +1 or -1 to indicate whether the flue gas flows from the first zone to the second zone or vice versa. The gap shape resistance correction factor is used to compensate for fluid friction loss in the flat channel of the door gap. This factor is a dimensionless empirical constant, and it is recommended to set it to 0.85 according to the standard fire door installation gap width characteristics. This is the current heat dissipation voltage value output by the wind speed probe; This is the base sustaining voltage under zero wind speed conditions; The heat exchange calibration coefficient is specific to the probe and its dimensions are strictly defined. The values ​​need to be obtained through wind tunnel calibration experiments, and they are related to the probe geometry and the fluid medium.

[0021] The above formula for calculating bidirectional wind speed is derived from the classic form of King's Law for hot-wire anemometers. , where n is usually taken as 0.5. Let Let the square of the base voltage be the value at zero wind speed. Solve This formula uses a quadratic relationship, which is a direct result of this mathematical transformation.

[0022] The difference between the first concentration value and the second concentration value is calculated and divided by the maximum value of the two to generate the concentration bias; the difference between the first temperature value and the second temperature value is calculated and divided by the maximum value of the two to generate the temperature bias. Preferably, the difference between the first concentration value and the second concentration value is calculated and divided by the maximum of the two to generate a concentration bias, including a protection step for outlier data: Extract the numerical pair consisting of the first concentration value and the second concentration value, and compare the maximum value of the two values ​​in the numerical pair with the preset minimum value; If the maximum value of both is less than the preset minimum value, the environment is determined to be in a smoke-free background state, and the concentration bias is directly assigned to zero to block the division operation. If the maximum value of the two values ​​is not less than the preset minimum value, perform a division operation step by dividing by the maximum value of the two values ​​to generate the concentration bias.

[0023] In this embodiment of the invention, a concentration bias is generated by calculating the difference between a first concentration value and a second concentration value and dividing by the maximum value of the two, and a temperature bias is generated by calculating the difference between a first temperature value and a second temperature value and dividing by the maximum value of the two. Specifically, a dynamic numerical pair consisting of the first and second concentration values ​​is extracted, and the relationship between the maximum value of the two values ​​and a preset minimum value is compared in real time. In this processing logic, the preset minimum value is typically set to a percentage of the background carbon monoxide concentration in the environment. to The adjustable range. If the numerical comparison determines that the maximum value of both is strictly less than the preset minimum value, the system directly determines that the current boundary space is in a completely smoke-free background state, directly assigns the concentration bias to zero, and forcibly blocks the division operation; conversely, if the maximum value of both is not less than the preset minimum value, it clearly indicates that at least one defense zone has been invaded by real fire smoke, and the system immediately activates and executes the extreme value normalization variant operation to generate concentration bias data. The extreme value normalization variant operation formulas for concentration bias and temperature bias are constructed as follows: In the above calculation formula, This represents the concentration bias of the output. and These are the first concentration values ​​of the first fire compartment and the second concentration values ​​of the second fire compartment, respectively, obtained from the analysis. For the temperature bias of synchronous output, and These are the first and second temperature values ​​on both sides, both in degrees Celsius.

[0024] Conventional relative deviation assessment models typically use a baseline structure of dividing the absolute difference between the two sides by the arithmetic mean of the two sides. However, in the extreme asymmetric condition of a fire on one side of the utility tunnel, the data on the fire side shows an exponential surge while the safe side maintains a low baseline value. The denominator of the conventional average value can cause nonlinear mutations or even overflows in the model output. The above-mentioned variant formula specially designed in this embodiment uses the maximum value function to directly lock the feature dimension most severely affected by the fire as the dynamic baseline. It not only strictly clamps the output of the bias within the linear constraint range of [-1,1], but also directly maps the positive and negative polarities of the values ​​to the relative physical location of the fire source. The absolute value accurately represents the tilt gradient of the smoke intrusion, which greatly improves the quantitative stability of the system for early asymmetric fires.

[0025] Calculate the sum of the first concentration value and the second concentration value, the sum of the first temperature value and the second temperature value, and the sum of the first visibility value and the second visibility value. Multiply each sum by its corresponding preset weight and sum them to generate the boundary fire spread index. Preferably, before generating the boundary fire spread index, the preset weights are dynamically mapped, including: The temperature change rate is generated by calculating the difference between the first temperature value in two adjacent sampling periods. Establish a preset weight adjustment matrix, which includes base concentration weight, base temperature weight, and base visibility weight. When the rate of temperature change is greater than the mutation threshold, the basic temperature weight is extracted from the preset weight adjustment matrix and multiplied by the preset amplification factor to generate the updated target temperature weight. The boundary fire spread index is generated by substituting the base concentration weight, target temperature weight, and base visibility weight as corresponding preset weights into the summation of each value.

[0026] In this embodiment of the invention, the underlying microprocessor extracts the first temperature value of the current sampling period. The first temperature value of the previous sampling period Divide the difference between the two by the time span of the adjacent sampling periods. Generating temperature change rate Then, a preset weight adjustment matrix is ​​created in memory and invoked. The matrix internally contains three dimensionless empirical basis factors, which are the basis concentration weights. Base temperature weighting With basic visibility weight Next, the real-time generated temperature change rate will be... It is compared with a preset mutation threshold, which is typically set to... The adjustable range is recommended to accurately identify the surge in space temperature caused by thermal convection during the pre-flame stage of a fire, while filtering out thermal disturbances caused by localized overheating of equipment. The baseline for judgment; when the rate of temperature change is strictly greater than the abrupt change threshold, the system extracts the base temperature weight based on the matrix index address. and multiply it by the preset magnification factor. To generate updated target temperature weights The amplification factor here is a dimensionless pure digital scalar. Its preset logic is based on the nonlinear surge characteristics of the environmental damage weight of high-temperature flue gas cross-regional spread in fluid thermodynamics. It is usually set to 1.2 to 1.8. In this embodiment, it is preferably set to 1.5, so as to give the temperature feature a higher weight in the fusion decision when the fire situation worsens.

[0027] After completing the weight mapping, the system calculates the sum of the first and second concentration values, the sum of the first and second temperature values, and the sum of the first and second visibility values, respectively. The variant fusion calculation formula for the boundary fire spread index is as follows: In this formula, BSI is the output boundary fire spread index; and These are the first concentration values ​​of the first fire compartment and the second concentration values ​​of the second fire compartment, respectively, obtained from the analysis. and These are the first and second temperature values ​​on both sides, respectively. and The visibility values ​​representing the smoke light reduction rate extracted from both sides are expressed as a percentage per meter. To achieve dimensionless measurement, three calibration reference characteristic constants are introduced into the denominator at the bottom of the formula, namely, the environmental concentration alarm characteristic constants. Ambient temperature alarm characteristic constant and ambient light reduction rate alarm characteristic constant The aforementioned baseline characteristic constants are all derived from the maximum alarm threshold parameter settings of the corresponding detectors in the national fire protection standards.

[0028] As another alternative implementation method, With temperature change rate Dynamic positive correlation, the specific relationship is as follows: in: The base magnification factor ranges from 1.0 to 1.2, with 1.0 being recommended. To adjust the slope, the value should be between 0.2 and 0.5, with 0.3 recommended. The mutation threshold is expressed in °C / s. To prevent Too small or too large a value loses its physical meaning; a lower limit should be set. upper limit That is, the final .

[0029] When using dynamic functional relationships, the more drastic the temperature change, the larger the amplification factor, thus giving the temperature characteristics a higher fusion decision weight when the fire situation deteriorates rapidly, thereby improving the sensitivity of early warning.

[0030] An adaptive control loop is triggered when the boundary fire spread index exceeds a safety threshold. The adaptive control loop includes: If the absolute values ​​of both concentration bias and temperature bias are less than the symmetry threshold, the dual-zone collaborative mode is triggered, and the highest frequency command is output to control the operation of the first and second exhaust fans. The deviation between the bidirectional wind speed value and the zero value is calculated, and the operating frequency of the first and second exhaust fans is dynamically adjusted synchronously according to the deviation until the bidirectional wind speed value converges to the target wind speed zone. Preferably, if the absolute values ​​of both concentration bias and temperature bias are less than the symmetry threshold, the dual-region cooperative mode is triggered, including: Using zero as the target value and bidirectional wind speed as the system feedback quantity, the deviation is extracted and substituted into the proportional-integral-differential equation. Extract the frequency correction increment from the output of the proportional-integral-differential equation; The frequency correction increment is added to the current frequency of the first blower to generate the first target adjustment frequency. The negative mapping value of the first target adjustment frequency is superimposed on the current frequency of the second blower to generate the second target adjustment frequency. The first target adjustment frequency is issued to the first blower, and the second target adjustment frequency is issued to the second blower.

[0031] If the absolute value of the concentration bias is greater than the one-way bias threshold, the one-sided defense mode is triggered. The positive or negative polarity of the concentration bias is determined to lock the fire zone. The output command controls the main exhaust fan of the fire zone to run at the highest frequency. The pressure difference deviation between the micro pressure difference value and the target smoke control pressure difference value is calculated. Based on the pressure difference deviation, the operating frequency of the adjacent supply fan in the non-fire zone is adjusted to establish a wind pressure barrier. Preferably, after determining the positive or negative polarity of the concentration bias to lock the ignition zone, dynamic relay smoke extraction is performed: If the concentration bias is positive, the first fire compartment is determined to be a fire compartment, and the second fire compartment is determined to be a non-fire compartment. Extract the second concentration value within the second fire compartment, and compare the second concentration value at the current time with that at the previous sampling time to generate the current climbing gradient; If the current climb gradient is greater than zero, a start command is generated and sent to the adjacent exhaust fan in the second fire compartment to perform a low-frequency cleaning task. If the current gradient is not greater than zero, output a shutdown command to smoothly shut down the adjacent exhaust fan.

[0032] Preferably, a start command is generated and sent to the adjacent exhaust fan corresponding to the second fire compartment to perform a low-frequency cleaning task, including: Extract the current operating frequency of adjacent blowers; Multiply the current operating frequency of the adjacent blowers by a preset safety ratio coefficient to calculate the upper limit of the exhaust frequency that will not damage the wind pressure barrier. The operating frequency of adjacent exhaust fans is clamped within the upper limit of the exhaust frequency to generate a frequency limit command, and the frequency limit command is sent to the frequency converter control terminal of the adjacent exhaust fans.

[0033] In this embodiment of the invention, when the controller embedded in the system determines in continuous time-series detection that the acquired boundary fire spread index is strictly greater than a set safety threshold, preferably a baseline of 1.2 representing a stable thermal plume established by the fire, the conventional ventilation standby state will be broken, and an adaptive control loop will be immediately triggered. This adaptive control loop first performs dynamic optimization and adaptive switching of the spatial fire distribution mode based on the bias characteristics generated by the preceding link.

[0034] The system first reads the absolute values ​​of concentration bias and temperature bias, and then rigorously compares them with a preset symmetry threshold. In this logic, the symmetry threshold is typically set to a decimal range of 0.1 to 0.3. To accurately capture the extreme symmetrical distribution of the fire source directly below the fire door, a baseline of 0.2 is recommended. If both the absolute values ​​of concentration bias and temperature bias are strictly less than this symmetry threshold, the system determines that the current fire source is spreading uniformly in both directions and triggers a dual-zone collaborative mode. In this mode, the system's underlying layer directly outputs the highest frequency command to force the first and second exhaust fans to operate at full load, aiming to create a globally lowest pressure trap at the interface. Simultaneously, to prevent the fans from blindly supplying air and causing smoke turbulence, the system uses zero as the target for absolute balance. The obtained bidirectional wind speed values ​​with polarity are used as system feedback, and the deviation is extracted and substituted into an improved discrete proportional-integral-differential equation to extract the frequency correction increment used for synchronously and dynamically adjusting the speed of the fans on both sides. This equation is constructed as follows: In this formula, The frequency correction increment for the current control cycle output; e(t) is the deviation between the currently acquired target zero value and the bidirectional wind speed value, in meters per second; The discrete sampling and adjustment period of the control system is expressed in seconds. This is the proportional gain coefficient, and its dimension is defined as Hertz per meter per second; Here is the integral gain coefficient, with dimensions of ; The differential gain coefficient has the following dimensions: .

[0035] The above uses an incremental PID control law, and the output is a frequency-corrected increment. It refers to, rather than, absolute frequency. Where: Proportional term: This corresponds to the proportional portion of a standard incremental PID controller. Integral term: , corresponding to the integral part; Differential term: , corresponding to the differential part.

[0036] In incremental PID, it is standard practice to use the second-order difference of the error (i.e., (e(t)-2e(t-1)+e(t-2))) for the derivative term to approximate the trend of the error rate of change, which is equivalent to the difference form of the derivative term in positional PID.

[0037] Calculation results Then, the system adds the current frequency of the first blower to generate the first target adjustment frequency. Simultaneously, based on the seesaw effect of the air pressure distribution at both ends of the enclosed corridor, the negative mapping value of this frequency correction increment is superimposed on the current frequency of the second blower to generate the second target adjustment frequency. Finally, the two complementary target adjustment frequencies are sent to the first and second blowers respectively. Verification under specific operating conditions: Assuming the deviation between the previous cycle and the two previous cycles is close to zero, and the currently read wind speed feedback is... (The wind is blowing from the first zone to the second zone), at this time the current deviation is... Set the control cycle to... , And ignore the extremely small differential terms. Substituting into the formula, we get... Assuming the original frequencies of both blowers are... Then the target frequency of the first blower changes to The target frequency of the second blower has been changed to The calculation results clearly indicate that when airflow flows back into the second zone, by reducing the air supply in the first zone and increasing the air supply in the second zone, a reverse pressure can be instantly established to force the wind speed to converge and push it back to the zero target zone.

[0038] Conversely, if the system determines that the absolute value of the concentration bias is strictly greater than the one-way bias threshold, it triggers the one-sided defense mode and determines the positive or negative polarity of the concentration bias to lock the fire zone. Assuming the bias is positive, the system determines the first fire zone as the fire zone and the second fire zone as the non-fire zone. At this time, the controller issues a command to control the main exhaust fan of the fire zone to run at the highest frequency and calculates the pressure difference deviation between the micro-pressure difference value at the gap and the target smoke control pressure difference value. A wind pressure barrier is established by adjusting the frequency of the second supply fan. To prevent a small amount of heat-leaked smoke from accumulating in the non-fire zone, the system simultaneously executes dynamic relay smoke exhaust control and extracts the second concentration value in the second fire zone. The second concentration value at the previous sampling time The difference is divided by the sampling period to obtain the current gradient. If the current gradient is strictly greater than zero, it indicates that secondary intrusion and leakage of flue gas is occurring, and the system immediately generates a start command and sends it to the adjacent exhaust fan to perform a low-frequency purging task. However, to avoid the suction effect of the adjacent exhaust fan from damaging the wind pressure barrier just established by the supply fan, the system introduces a damage-prevention clamping variant calculation: extracting the current operating frequency of the second supply fan. Multiply it by the preset safety ratio coefficient Calculate the upper limit of exhaust frequency without damaging the wind pressure barrier. This safety ratio factor is a dimensionless constant, and based on the shaft power matching characteristics of the supply and exhaust fans, it is typically set between 0.3 and 0.6, preferably 0.4 to ensure that the exhaust volume is always significantly inferior. Assuming the frequency adjustment of the second supply fan required for current pressure build-up is... The system calculates that the upper limit of the exhaust frequency is The system will then calculate and generate The frequency limit command is sent to the frequency converter control terminal of the adjacent exhaust fan.

[0039] Within the adaptive control cycle, the time-varying derivative of the micro-pressure difference is calculated. If the time-varying derivative is greater than the airflow oscillation threshold, the damping lockout procedure is triggered, the calculation of the new adjustment frequency of each blower and exhaust fan is stopped, and the equipment is frozen at the current operating frequency until the damping countdown ends.

[0040] Preferably, calculating the time-varying derivative of the micro-pressure difference value within the adaptive control cycle includes: Construct a data queue with a predetermined number of bits, and push the micro pressure difference values ​​generated during the adaptive control cycle into the data queue point by point; Extract the difference between the latest micro-pressure differential value and the last micro-pressure differential value in the data queue, and divide the difference by the time span covered by the data queue to generate the time variation derivative; After the damping countdown ends, the historical cache inside the data queue is cleared, and the current micro-pressure difference value is pushed in again to activate the adaptive control loop.

[0041] Preferably, after triggering the adaptive control loop, the system's embedded controller also performs the following exit decision step: Extract the boundary fire spread index calculated for each sampling period and construct a status monitoring array based on the time series; Iterate through the values ​​in the state monitoring array to calculate the difference sequence of adjacent elements, and detect consecutive counts where all elements in the difference sequence are less than zero. If the continuous count value reaches the preset frame number and the latest boundary fire spread index is lower than the safety baseline, it is determined that the fire in the boundary area has been eliminated, and the adaptive control loop is terminated.

[0042] In this embodiment of the invention, during the high-frequency execution cycle of the adaptive control loop, to prevent the frequent large-span frequency adjustments of the supply and exhaust fans from superimposing with the airflow turbulence at the gaps in the smokeproof doors, thus causing low-frequency airflow resonance in the pipeline system, the system first constructs a data queue with a predetermined number of bits in memory, and then pushes the micro-pressure difference values ​​collected and analyzed in real time during the adaptive control loop into this data queue point by point. To avoid occupying too much memory stack while ensuring coverage of a complete typical airflow pulsation cycle, a predetermined number of bits is recommended to be preferably 20 storage points. Whenever new data is pushed onto the stack, the system immediately extracts the difference between the latest pushed micro-pressure difference value in the data queue and the micro-pressure difference value about to be squeezed out at the end of the queue. Given that the traditional differential derivative calculated solely based on two adjacent sampling points is easily submerged by the inherent high-frequency white noise of the micro-pressure difference sensor, leading to misjudgment of the derivative's polarity and magnitude, this embodiment introduces a modified calculation formula for the macroscopic time-varying derivative based on the time window span: In this formula, The final time-varying derivative characterizes the intensity of transient pressure fluctuations at the interface. This is the latest differential pressure value pushed into the data queue at the current moment, in Pascals; This represents the historical micro-pressure differential value at the very end of the data queue, with the unit being Pascals. The predetermined number of bits set for the data queue is a dimensionless count value; This represents the discrete time span of a single sampling of the underlying controller, in seconds.

[0043] Next, the absolute value of the calculated time-varying derivative is compared with a preset airflow oscillation threshold, the unit of which is... Taking a specific working condition as an example, assume the current single sampling span is... The queue capacity is set to 20, and the latest pressure difference is extracted as follows: The final historical pressure difference is (This indicates a dramatic reversal of airflow direction across regions), substituting into the formula to calculate... The value did not exceed The system continues to maintain normal regulation when the threshold is reached; however, if the absolute value calculated under extreme conditions is as high as [a certain threshold], the system will continue to maintain normal regulation. If the airflow oscillation threshold is exceeded, the system immediately triggers the damping lockout procedure, forcibly stopping the calculation of the proportional-integral-derivative (PID) adjustment frequencies of each blower and exhaust fan. All frequency converters are frozen at their current operating frequencies until the damping countdown ends. After the countdown, all historical buffers in the data queue are cleared, and the current stable micro-pressure difference is re-injected to reactivate the adaptive control loop. Simultaneously, to prevent the system from falling into an endless smoke exhaust loop after the fire is extinguished, the controller also executes an exit decision step: in each sampling cycle, the boundary fire spread index calculated by the front-end link (which is dimensionless due to normalization in the front-end calculation) is extracted in real time and stored frame by frame in the system's state monitoring array according to the time series. The system iterates through this state monitoring array each cycle, calculates the difference sequence of elements at adjacent time nodes, and sets up a dedicated accumulation register to strictly detect continuous counts where all elements in the difference sequence are strictly less than zero (indicating an irreversible monotonically decaying fire trend). If the continuous count value is corrupted (i.e., a bounce difference greater than or equal to zero occurs), the accumulation register is immediately cleared and recalculated. Only when the continuous count value reaches the preset number of frames and the latest extracted boundary fire spread index is strictly lower than the system's set safety baseline (which represents the environment returning to a background state containing only slight residual dust) will the system finally determine that the fire in the boundary area has been completely eliminated, and then safely terminate and exit the adaptive control loop, and then enter the downgraded low-frequency cruise ventilation mode.

[0044] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0045] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0046] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

[0047] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A fire emergency ventilation and smoke exhaust system for a utility tunnel, comprising a first fire compartment, a second fire compartment, and a sensor array and air supply and exhaust equipment disposed at the boundary between the two compartments, characterized in that, The embedded controller performs the following smoke extraction coordination steps: Acquire the first concentration value, the second concentration value, the first temperature value, the second temperature value, the first visibility value, the second visibility value, the micro-pressure difference value and the bidirectional wind speed value at the junction gap on both sides; The difference between the first concentration value and the second concentration value is calculated and divided by the maximum value of the two to generate the concentration bias; the difference between the first temperature value and the second temperature value is calculated and divided by the maximum value of the two to generate the temperature bias. Calculate the sum of the first concentration value and the second concentration value, the sum of the first temperature value and the second temperature value, and the sum of the first visibility value and the second visibility value. Multiply each sum by its corresponding preset weight and sum them to generate the boundary fire spread index. An adaptive control loop is triggered when the boundary fire spread index exceeds a safety threshold. The adaptive control loop includes: If the absolute values ​​of both concentration bias and temperature bias are less than the symmetry threshold, the dual-zone collaborative mode is triggered, and the highest frequency command is output to control the operation of the first and second exhaust fans. The deviation between the bidirectional wind speed value and the zero value is calculated, and the operating frequency of the first and second exhaust fans is dynamically adjusted synchronously according to the deviation until the bidirectional wind speed value converges to the target wind speed zone. If the absolute value of the concentration bias is greater than the one-way bias threshold, the one-sided defense mode is triggered. The positive or negative polarity of the concentration bias is determined to lock the fire zone. The output command controls the main exhaust fan of the fire zone to run at the highest frequency. The pressure difference deviation between the micro pressure difference value and the target smoke control pressure difference value is calculated. Based on the pressure difference deviation, the operating frequency of the adjacent supply fan in the non-fire zone is adjusted to establish a wind pressure barrier. Within the adaptive control cycle, the time-varying derivative of the micro-pressure difference is calculated. If the time-varying derivative is greater than the airflow oscillation threshold, the damping lockout procedure is triggered, the calculation of the new adjustment frequency of each blower and exhaust fan is stopped, and the equipment is frozen at the current operating frequency until the damping countdown ends.

2. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 1, characterized in that, Acquire the first concentration value, second concentration value, first temperature value, second temperature value, first visibility value, second visibility value, as well as the micro-pressure difference value and bidirectional wind speed value at the boundary gap on both sides, including: Multimodal raw sensor signals from both sides and gaps at the interface are collected synchronously according to the preset sampling frequency, and the multimodal raw sensor signals are time-series aligned according to time points to construct a time-synchronized dataset. The first environmental electrical signal corresponding to the first fire compartment, the second environmental electrical signal corresponding to the second fire compartment, and the gap fluid electrical signal corresponding to the boundary gap are separated from the time synchronization dataset. Extract the numerical sets of the first environmental electrical signal, the second environmental electrical signal, and the gap fluid electrical signal within the sliding time window and calculate the average value to generate the first smoothed signal, the second smoothed signal, and the gap smoothed signal after filtering out high-frequency noise, respectively. The first smoothed signal is calibrated and mapped using physical quantities to obtain the first concentration value, the first temperature value, and the first visibility value. The second smoothed signal is calibrated and mapped using physical quantities to obtain the second concentration value, the second temperature value, and the second visibility value. The environmental pressure difference feature and airflow movement feature are extracted by decoupling the gap smoothing signal. The micro pressure difference value is calculated based on the environmental pressure difference feature, and the airflow movement feature is combined with the airflow movement direction factor to generate bidirectional wind speed value through vector calculation.

3. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 1, characterized in that, Before generating the boundary fire spread index, the preset weights are dynamically mapped, including: The temperature change rate is generated by calculating the difference between the first temperature value in two adjacent sampling periods. Establish a preset weight adjustment matrix, which includes base concentration weight, base temperature weight, and base visibility weight. When the rate of temperature change is greater than the mutation threshold, the basic temperature weight is extracted from the preset weight adjustment matrix and multiplied by the preset amplification factor to generate the updated target temperature weight. The boundary fire spread index is generated by substituting the base concentration weight, target temperature weight, and base visibility weight as corresponding preset weights into the summation of each value.

4. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 1, characterized in that, If the absolute values ​​of both concentration bias and temperature bias are less than the symmetry threshold, the dual-region cooperative mode is triggered, including: Using zero as the target value and bidirectional wind speed as the system feedback quantity, the deviation is extracted and substituted into the proportional-integral-differential equation. Extract the frequency correction increment from the output of the proportional-integral-differential equation; The frequency correction increment is added to the current frequency of the first blower to generate the first target adjustment frequency. The negative mapping value of the first target adjustment frequency is superimposed on the current frequency of the second blower to generate the second target adjustment frequency. The first target adjustment frequency is issued to the first blower, and the second target adjustment frequency is issued to the second blower.

5. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 1, characterized in that, After determining the positive or negative polarity of the concentration bias to lock the ignition zone, dynamic relay smoke extraction is then implemented: If the concentration bias is positive, the first fire compartment is determined to be a fire compartment, and the second fire compartment is determined to be a non-fire compartment. Extract the second concentration value within the second fire compartment, and compare the second concentration value at the current time with that at the previous sampling time to generate the current climbing gradient; If the current climb gradient is greater than zero, a start command is generated and sent to the adjacent exhaust fan in the second fire compartment to perform a low-frequency cleaning task. If the current gradient is not greater than zero, output a shutdown command to smoothly shut down the adjacent exhaust fan.

6. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 5, characterized in that, A start command is generated and sent to the adjacent exhaust fan corresponding to the second fire compartment to perform a low-frequency cleaning task, including: Extract the current operating frequency of adjacent blowers; Multiply the current operating frequency of the adjacent blowers by a preset safety ratio coefficient to calculate the upper limit of the exhaust frequency that will not damage the wind pressure barrier. The operating frequency of adjacent exhaust fans is clamped within the upper limit of the exhaust frequency to generate a frequency limit command, and the frequency limit command is sent to the frequency converter control terminal of the adjacent exhaust fans.

7. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 1, characterized in that, Calculate the time-varying derivative of the micro-pressure difference value within the adaptive control cycle, including: Construct a data queue with a predetermined number of bits, and push the micro pressure difference values ​​generated during the adaptive control cycle into the data queue point by point; Extract the difference between the latest micro-pressure differential value and the last micro-pressure differential value in the data queue, and divide the difference by the time span covered by the data queue to generate the time variation derivative; After the damping countdown ends, the historical cache inside the data queue is cleared, and the current micro-pressure difference value is pushed in again to activate the adaptive control loop.

8. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 1, characterized in that, The concentration bias is calculated by calculating the difference between the first and second concentration values ​​and dividing by the maximum of the two values, including protection steps for outlier data. Extract the numerical pair consisting of the first concentration value and the second concentration value, and compare the maximum value of the two values ​​in the numerical pair with the preset minimum value; If the maximum value of both is less than the preset minimum value, the environment is determined to be in a smoke-free background state, and the concentration bias is directly assigned to zero to block the division operation. If the maximum value of the two values ​​is not less than the preset minimum value, perform a division operation step by dividing by the maximum value of the two values ​​to generate the concentration bias.

9. The emergency ventilation and smoke exhaust system for pipe gallery fires as described in claim 1, characterized in that, After triggering the adaptive control loop, the system's embedded controller also performs the following exit decision step: Extract the boundary fire spread index calculated for each sampling period and construct a status monitoring array based on the time series; Iterate through the values ​​in the state monitoring array to calculate the difference sequence of adjacent elements, and detect consecutive counts where all elements in the difference sequence are less than zero. If the continuous count value reaches the preset frame number and the latest boundary fire spread index is lower than the safety baseline, it is determined that the fire in the boundary area has been eliminated, and the adaptive control loop is terminated.

Images