Intelligent control method for tunnel fire smoke exhaust fan based on physical priori and adaptive PID

Abstract

This invention discloses an intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID. The method first constructs a numerical model of the tunnel fire using an FDS (Functional Data System) and establishes a Python-FDS bidirectional coupling interface. It then establishes a physical prior system for tunnel fire smoke exhaust control, forming physical boundary constraints and tuning prior rules. A parameter-adaptive PID controller integrating differential priors, online parameter adaptation throughout the fire cycle, and integral anti-saturation is designed and its output is corrected using physical boundary constraints. A data preprocessing and anti-interference compensation module for measurement points is constructed. Finally, a hybrid control objective function integrating control error loss, physical prior constraints, and airflow smoothing constraints is built, and parameter verification is completed to achieve real-time closed-loop intelligent control of fan volumetric flow rate. This invention can shorten adjustment time, reduce overshoot, and improve control stability and physical rationality, providing technical support for tunnel fire smoke control, personnel evacuation, and emergency rescue.

Description

Technical Field

[0001] This invention relates to the interdisciplinary fields of civil engineering disaster prevention and mitigation, tunnel fire protection engineering and intelligent control, and in particular to an intelligent control method for tunnel fire smoke exhaust fans based on physical prior and adaptive PID. It can be applied to intelligent monitoring and control systems for tunnel fires, smoke exhaust auxiliary decision-making systems and fire emergency rescue platforms, to achieve adaptive, highly stable and physically reasonable intelligent control of the volumetric flow rate of smoke exhaust fans during all times of tunnel fires. Background Technology

[0002] The long, narrow, and enclosed spatial structure of tunnels causes hot, toxic smoke to spread rapidly longitudinally after a fire breaks out. When longitudinal ventilation is insufficient, the hot smoke can spread upstream of the fire source, forming a smoke backflow zone. This seriously threatens the evacuation safety of personnel upstream of the fire source and the access route for firefighters, making it the primary cause of casualties in tunnel fires. Precise and rapid control of smoke exhaust fans is the core means to suppress the spread of smoke in tunnel fires and ensure the safety of personnel evacuation.

[0003] Currently, the methods for controlling smoke exhaust fans in tunnel fires are mainly divided into three categories: The first category is the fixed air volume ventilation mode, which presets a fixed fan volume flow rate based on the design conditions. This method cannot adapt to the dynamic increase of fire source power and the spatiotemporal changes in smoke spread during the fire development process. It is prone to insufficient air volume leading to smoke backflow, or excessive air volume leading to turbulent smoke diffusion, resulting in extremely poor fire control effect. The second category is the conventional fixed parameter PID control method, which achieves closed-loop control of fan volume flow rate through single-point temperature feedback. Although this method can achieve dynamic adjustment, tunnel fire systems have strong nonlinearity, large lag, and strong time-varying characteristics. The dynamic characteristics of different stages of fire development vary greatly, and fixed parameter PID cannot achieve optimal control at all times, which is generally not feasible during fires. The problems of excessive initial control, large overshoot, delayed adjustment response, and large fluctuations in the steady-state stage make it difficult to meet the rapid and stable control requirements in emergency scenarios. The third category is intelligent control methods, such as fuzzy control and model predictive control. Although these methods can improve control performance, they generally have two core defects: First, they lack specific physical constraints on the flow and ventilation characteristics of smoke in tunnel fires, which can easily lead to control outputs that do not conform to engineering realities, such as reverse ventilation and air volume exceeding the rated operating conditions of the fan. Their physical rationality is insufficient, making them difficult to implement. Second, the coupling between existing control algorithms and CFD numerical simulations is mostly in offline simulation mode, which cannot achieve real-time synchronous iteration between the control algorithm and the dynamic evolution of the fire. The control effect is seriously out of sync with the actual engineering scenario. Summary of the Invention

[0004] To address the aforementioned technical shortcomings, the purpose of this invention is to provide an intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID. This method can significantly improve the stability and control accuracy of tunnel fire smoke exhaust control throughout the entire time period while maintaining millisecond-level real-time response speed, providing reliable core control technology support for tunnel fire smoke prevention and emergency rescue.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0006] This invention provides an intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID, comprising the following steps:

[0007] Step 1: Construct a tunnel fire FDS numerical model with dynamic ventilation boundaries, and build a Python-FDS bidirectional coupling interface between the tunnel fire FDS numerical model and the control algorithm to achieve synchronous iteration of CFD numerical calculation and control algorithm.

[0008] Step 2: Construct a physical prior system for smoke exhaust control in tunnel fires, generate physical boundary constraints to constrain the output of the fan control, and tuning prior rules to limit the parameters of the adaptive PID controller.

[0009] Step 3: Design a parameter adaptive PID controller. The parameter adaptive PID controller integrates a derivative-first module, a fire-fighting all-time parameter online adaptive module, and an integral anti-saturation module, and uses the physical boundary constraints to correct the output of the parameter adaptive PID controller.

[0010] Step 4: Construct a preprocessing and anti-interference compensation module for the measurement point data, which is used to denoise and process outliers in the real-time monitoring data of the temperature sensor, and provide a stable feedback signal for the parameter adaptive PID controller;

[0011] Step 5: Construct a hybrid control objective function that integrates control error loss, physical prior constraints, and air volume smoothing constraints, and complete the parameter tuning and multi-condition simulation verification of the parameter adaptive PID controller based on this hybrid control objective function;

[0012] Step 6: Based on the tuned parameter adaptive PID controller, the real-time closed-loop intelligent control of the volumetric flow rate of the tunnel fire smoke exhaust fan is realized through the Python-FDS bidirectional coupling interface.

[0013] Preferably, in step 1, the construction of the FDS numerical model for tunnel fires and the establishment of the Python-FDS bidirectional coupling interface specifically include:

[0014] Step 1.1: Establish a tunnel fire FDS numerical model that matches the actual engineering situation, and set up tunnel fire simulation scenarios with multiple fire source locations, multiple fire source power levels, and multiple longitudinal ventilation wind speed levels;

[0015] Step 1.2: Adopt a dynamic grid division strategy, using the first grid size in the area near the fire source and fan, and the standard grid in the remaining areas; the first grid size is smaller than the standard grid size;

[0016] Step 1.3: Configure the fan and define its entire air supply section as the VENT boundary, set its initial volumetric flow rate to the initial shutdown state, and set upper and lower limits for the fan's volumetric flow rate as hard constraints.

[0017] Step 1.4: Install multiple temperature sensors at fixed intervals along the longitudinal direction of the tunnel, and set up the main feedback measuring point on the roof upstream of the fire source, while setting up multiple backup measuring points upstream and downstream of the fire source.

[0018] Step 1.5: Synchronously execute the following Python script at each calculation time step: Extract real-time temperature data from the temperature sensor from the FDS numerical model of the tunnel fire, input it into the parameter adaptive PID controller, and update the target fan volume flow rate output by the parameter adaptive PID controller to the flow parameters at the VENT boundary in real time, so as to achieve continuous control that is completely synchronized with the FDS calculation step.

[0019] Preferably, in step 2, the physical prior system includes the prior length of flue gas propagation, physical boundary constraints for airflow control, and tuning prior rules; wherein,

[0020] The physical prior for the smoke propagation length is calculated by combining the critical temperature rise threshold method and the thermal center of mass method; the generated prior length L of the smoke propagation upstream of the fire source is... phy (t), the calculation formula is:

[0021] ;

[0022] In the formula, L coarse (t) represents the coarse location length of the flue gas front calculated using the critical temperature rise threshold method, L cm (t) represents the flue gas propagation length calculated using the thermal center-of-mass method, w coarse For the weight of the thick front, w cm The weight of the thermal centroid is w. coarse +w cm =1;

[0023] The physical boundary constraints for airflow control generate dynamic upper and lower limits for fan volumetric flow rate based on the prior length of smoke propagation, fire source power, and tunnel ventilation characteristics. The calculation formula is as follows:

[0024] ;

[0025] ;

[0026] In the formula, V rated Q(t) represents the rated maximum volumetric flow rate of the fan; Q(t) represents the power of the fire source at time t; k1 and k2 are the tunnel ventilation characteristic coefficients, which are determined by multi-condition simulation fitting.

[0027] The aforementioned prior rules are based on the full-time evolution characteristics of tunnel fires, dividing the fire process into an initial growth period, a dynamic development period, and a quasi-steady-state control period, and setting prior intervals for the proportional coefficient, integral coefficient, and differential coefficient of each stage.

[0028] Preferably, in step 3, designing the parameter adaptive PID controller specifically includes:

[0029] Step 3.1: Construct a basic closed-loop PID control system, using the real-time temperature T at the core control measurement point upstream of the fire source. d (t) and target control temperature T t The difference is taken as the control error e(t), and the target control temperature T is set. t The error calculation formula is:

[0030] ;

[0031] To address the discrete-time step computation requirements of FDS coupled with Python scripts, the above equation is discretized to obtain a discrete form that can be directly used for code iteration, as shown in the following expression:

[0032] ;

[0033] In the formula, t is the current calculation time step number, u(t) is the output control law of the parameter adaptive PID controller, and K p (t), K i (t) and K d (t) represents the proportional, integral, and differential coefficients of the current step adaptive tuning;

[0034] Step 3.2: Design the differential-first module to address the overshoot problem caused by the sudden temperature rise in the early stages of a fire. The differential element is applied to the temperature feedback signal rather than the error signal to avoid drastic fluctuations in the control quantity caused by sudden changes in the setpoint. The differential-first output calculation formula is as follows:

[0035] ;

[0036] In the formula, u d (t) represents the output of the differential lead-in stage;

[0037] Step 3.3: Design an online adaptive module for all-time fire parameters. Based on fuzzy inference rules, the module uses the control error e(t) and the error change rate Δe(t) = e(t) - e(t-1) as inputs, and the adaptive adjustment amount ΔK of the PID parameters as inputs. p ΔK i ΔK d For the output, the PID parameters are tuned online in real time, taking into account the prior parameter range in step 2.

[0038] Step 3.4: Design an integral anti-saturation module, which combines integral limiting and back-calculation compensation. When the control output reaches the upper and lower limits of the air volume constraint, the accumulation of the integral term is stopped. At the same time, the cumulative error of integral saturation is eliminated by back calculation to avoid system oscillation and overshoot caused by excessive accumulation of the integral term.

[0039] Preferably, the target control temperature T is set. t If the temperature is 100℃, then the specific method for real-time online tuning of PID parameters is as follows:

[0040] (1) When |e(t)|>80℃, increase K p Decrease K i It can quickly suppress temperature deviations and avoid over-control.

[0041] (2) When 30℃ < |e(t)| ≤ 80℃, the balance adjustment K p K i K d It balances adjustment speed and stability;

[0042] (3) When |e(t)|≤30℃, decrease K p Increase K i This improves the accuracy of steady-state control and suppresses fluctuations.

[0043] Preferably, step 4, constructing the preprocessing and anti-interference compensation module for the measurement point data, includes the following steps:

[0044] Step 4.1: Denoise the temperature data collected in real time by the temperature sensor using a first-order exponential moving average method. The calculation formula is as follows:

[0045] ;

[0046] In the formula, Tˊ d (t) represents the denoised temperature value, and α is the smoothing coefficient, with a value ranging from 0.2 to 0.5;

[0047] Step 4.2: Identify and process outliers in the denoised temperature data. Determine outliers based on the 3σ criterion. For outliers that exceed the physically reasonable range, replace them with the average of the previous and next 3 frames of data. For more than 3 consecutive frames of outlier data, automatically switch to the temperature data of the backup measuring point to achieve fault tolerance of the temperature sensor.

[0048] Step 4.3: Design an anti-interference compensation module for sudden changes in operating conditions. Based on the temperature gradient change rate upstream and downstream of the fire source, identify sudden changes in operating conditions such as a sudden increase in fire source power and flue gas backflow. When a sudden change in operating conditions is detected, automatically trigger the rapid adjustment mechanism of PID parameters, increase the weight of the proportional coefficient, shorten the system response time, and improve the anti-interference capability of the parameter adaptive PID controller.

[0049] Preferably, in step 5, the construction and parameter tuning of the hybrid control objective function specifically includes:

[0050] Step 5.1: Construct the control error loss term, and use the sum of squared errors index to measure control accuracy. The calculation formula is as follows:

[0051] ;

[0052] In the formula, N is the total simulation time steps;

[0053] Step 5.2: Construct a physical prior constraint loss term to penalize control quantities that exceed the physical prior boundary. The calculation formula is as follows:

[0054] ;

[0055] In the formula, V(n) is the volumetric flow rate control value of the fan at time t, and λ phy This is the physical constraint weighting coefficient, with a value ranging from 0.5 to 1.0;

[0056] Step 5.3: Construct a smoothing loss term for airflow to suppress drastic fluctuations in the fan's volumetric flow rate and avoid frequent start-ups and shutdowns and equipment damage. The calculation formula is as follows:

[0057] ;

[0058] In the formula, λ smooth The smoothing constraint weight coefficients are set to a value between 0.001 and 0.01.

[0059] Step 5.4: The mixed control objective function is a weighted sum of the loss terms, calculated using the following formula:

[0060] ;

[0061] Step 5.5: Based on multi-condition simulation data, with the minimization of the hybrid control objective function as the optimization objective, complete the initial tuning of the PID parameters, determine the initial values ​​of the parameters at each stage and the membership function of the fuzzy inference rule; through batch simulations of various fire source locations, various fire source powers and various wind speed conditions, complete the generalization verification and parameter optimization of the parameter adaptive PID controller.

[0062] Preferably, in step 6, the real-time closed-loop intelligent control specifically includes:

[0063] Step 6.1: Load the tuned parameters of the adaptive PID controller, physical prior constraint parameters, data preprocessing smoothing coefficients, and standardized parameters;

[0064] Step 6.2: Extract real-time monitoring data from the temperature sensor in the tunnel step by step through the Python-FDS bidirectional coupling interface. After noise reduction preprocessing and outlier processing, input the data into the parameter adaptive PID controller.

[0065] Step 6.3: The parameter adaptive PID controller completes the online adaptive tuning of PID parameters based on the control error and error change rate at the current moment, calculates the output control rate, and converts it into the target volume flow rate of the fan at the next moment after correction by physical prior boundary constraints.

[0066] Step 6.4: Update the target volumetric flow rate to the VENT boundary of the FDS numerical model for tunnel fire in real time through the Python-FDS bidirectional coupling interface, complete the CFD calculation and control iteration for the next time step, and realize the real-time closed-loop control of the fan volumetric flow rate.

[0067] Step 6.5: Real-time storage of temperature timing data at measurement points, fan volume flow control commands, and PID parameter adaptive adjustment data during the control process; generation of control effect analysis reports and visualization curves; and synchronous output to the tunnel fire emergency prevention and control platform.

[0068] Beneficial effects:

[0069] 1. This invention deeply integrates the hybrid physical prior knowledge specific to tunnel fires with the parameter adaptive PID controller, effectively overcoming the shortcomings of traditional fixed parameter PID controllers in tunnel fires with strong time-varying, large lag, and strong nonlinear conditions, such as over-control, lag, and large steady-state fluctuations. Through the synergistic effect of derivative-first, online adaptive tuning, and integral anti-saturation, it shortens the smoke exhaust control adjustment time, reduces overshoot, and significantly improves control response speed, adjustment accuracy, and full-cycle operation stability.

[0070] 2. This invention achieves synchronous iteration of the control algorithm and fire evolution through real-time interaction via Python-FDS bidirectional coupling. Combined with noise reduction of measurement point data, fault tolerance for outliers, anti-interference mechanism for sudden changes in operating conditions, and air volume smoothing constraint mechanism, it ensures that the output of the PID controller strictly conforms to the flue gas flow law and the rated operating range of the fan. This fundamentally avoids unreasonable phenomena such as reverse ventilation, over-range output, and frequent and drastic equipment adjustments, and greatly improves the robustness, physical rationality, and engineering applicability of the system.

[0071] 3. This invention can perform real-time, precise, and closed-loop intelligent control of the volumetric flow rate of smoke exhaust fans in tunnel fires, effectively suppressing the backflow and longitudinal spread of high-temperature toxic smoke, stabilizing the safety of personnel evacuation and fire rescue channels upstream of the fire source, providing stable and reliable core technical support for emergency monitoring, intelligent prevention and control, and rescue decision-making in tunnel fires, and significantly improving the comprehensive emergency response capability and safety assurance level of tunnel fires. Attached Figure Description

[0072] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0073] Figure 1 A flowchart of the intelligent control method for tunnel fire smoke exhaust fans based on physical prior and adaptive PID provided in an embodiment of the present invention;

[0074] Figure 2 A schematic diagram of the FDS numerical model of a tunnel fire with dynamic ventilation boundaries provided in an embodiment of the present invention;

[0075] Figure 3 This is a flowchart of the Python-FDS bidirectional coupled solution process provided in an embodiment of the present invention;

[0076] Figure 4 The PID temperature and fan volume flow rate control curve provided in the embodiment of the present invention. Detailed Implementation

[0077] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0078] like Figure 1 , Figure 3As shown, this embodiment of the invention provides an intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID, including the following steps:

[0079] Step 1: Construct a numerical model of tunnel fire FDS with dynamic ventilation boundaries, and establish a Python-FDS bidirectional coupling interface between the tunnel fire FDS numerical model and the control algorithm to achieve synchronous iteration of CFD numerical calculation and control algorithm; specific steps include:

[0080] Step 1.1: Establish a numerical model of tunnel fire using FDS (Fire Distance Analysis) to match the actual engineering situation. The tunnel dimensions are set as 300m long, 9m wide, and 6m high. Five fire source locations are set at distances of 50m, 100m, 150m, 200m, and 250m from the tunnel entrance. The fire source power is set in 8 levels, ranging from 5MW to 40MW in 5MW intervals. The longitudinal ventilation velocity is set in 12 levels, ranging from 1m / s to 12m / s in 1m / s intervals. (See [link / reference]) Figure 2 ;

[0081] The fire source is set as a rectangular area with a cross-section of 2m×5m, 0.5m above the ground. Its power growth follows the t² fire model, and the total simulation time is no less than 600s. The fuel type is heptane, and the carbon monoxide generation rate and smoke particle production rate in its combustion products are set to 0.006 and 0.015, respectively. The tunnel wall material is defined as concrete, and its thermal parameters are set using the default configuration of FDS software.

[0082] Step 1.2: A dynamic mesh generation strategy is adopted. The first mesh size is 0.25m×0.25m×0.25m, with a radius of 50m from the center of the fire source and within 5m around the fan. The remaining area of ​​the tunnel uses a standard mesh of 0.5m×0.5m×0.5m. The boundary conditions at both ends of the tunnel are set to "OPEN".

[0083] Step 1.3: The ventilation system consists of two sets of fans, symmetrically arranged at distances of 75m and 225m from the tunnel entrance. Each set contains two fans. The dimensions of a single fan are 3m × 1.5m × 1.5m. The top of the fan is installed 1m from the tunnel ceiling and is equipped with a 1m diameter air supply section. The duct and fan components of the HVAC system in the tunnel fire FDS numerical model are removed. The entire air supply section of the fan is defined as the VENT boundary, and the initial volumetric flow rate is set to 0m³ / s to match the initial shutdown state of the fan.

[0084] Step 1.4: Install temperature sensors at 5m intervals along the longitudinal centerline of the tunnel, with a total of 61 temperature sensors and a sampling time interval of 1s; set up a control measuring point on the ceiling 10m upstream of the fire source as the main feedback signal source for the PID controller; set up 3 backup monitoring measuring points upstream and downstream of the fire source for fault tolerance of temperature sensors and auxiliary judgment of fire situation.

[0085] Step 1.5: Execute synchronously in each calculation time step using a Python script: Extract real-time temperature data from the temperature sensor from the FDS numerical model of the tunnel fire and input it to the parameter adaptive PID controller. Update the target fan volumetric flow rate (also known as air volume) output by the parameter adaptive PID controller to the VOLUMEFLOW parameter at the VENT boundary in real time, achieving continuous control fully synchronized with the FDS calculation step. Simultaneously, set hard constraints on the upper and lower limits of the flow rate (0~40 m³ / s) in the Python script to match the rated operating range of the fan.

[0086] Step 2: Construct a physical prior system for smoke exhaust control in tunnel fires, generate physical boundary constraints to constrain the output of the fan control, and tuning prior rules to limit the parameters of the adaptive PID controller.

[0087] The physical prior for the smoke propagation length is calculated by combining the critical temperature rise threshold method and the thermal center of mass method; the generated prior length L of the smoke propagation upstream of the fire source is... phy (t), the calculation formula is:

[0088] ;

[0089] In the formula, L coarse (t) represents the coarse location length of the flue gas front calculated using the critical temperature rise threshold method, L cm (t) represents the flue gas propagation length calculated using the thermal center-of-mass method, w coarse For the weight of the thick front, w cm The weight of the thermal centroid is w. coarse +w cm =1;

[0090] The physical boundary constraints for airflow control generate dynamic upper and lower limits for fan volumetric flow rate based on the prior length of smoke propagation, fire source power, and tunnel ventilation characteristics. The calculation formula is as follows:

[0091] ;

[0092] ;

[0093] In the formula, V ratedQ(t) represents the rated maximum volumetric flow rate of the fan; Q(t) represents the power of the fire source at time t; k1 and k2 are the tunnel ventilation characteristic coefficients, which are determined by multi-condition simulation fitting.

[0094] The tuning prior rules are based on the full-time evolution characteristics of tunnel fires, dividing the fire process into three stages: the initial growth stage (0~120s), the dynamic development stage (120~300s), and the quasi-steady-state control stage (300~600s). Prior intervals for PID parameter tuning are defined for each stage.

[0095] (1) Initial growth period: proportional coefficient K p The prior interval is [0.002, 0.008], and the integral coefficient K i The prior interval is [-0.002, 0], and the differential coefficient K d The prior interval is [-0.0001, 0], which strengthens the differential advance to suppress overshoot;

[0096] (2) Dynamic development period: proportional coefficient K p The prior interval is [0.001, 0.005], and the integral coefficient K i The prior interval is [-0.0015, -0.0005], and the differential coefficient K... d The prior interval is [-0.00005, 0], balancing the adjustment speed and stability;

[0097] (3) Quasi-steady-state control period: The prior interval of the proportional coefficient Kp is [0.0005, 0.003], the prior interval of the integral coefficient Ki is [-0.001, -0.0002], and the prior interval of the differential coefficient Kd is [-0.00002, 0], to enhance the integral anti-saturation and steady-state accuracy;

[0098] ;

[0099] In the formula, u(t) is the output parameter of the PID controller; e(t) is the control error, see step 3.1 for details.

[0100] Step 3: See Figure 3 A parameter adaptive PID controller is designed, which integrates a derivative-first module, a fire-fighting all-time parameter online adaptive module, and an integral anti-saturation module, and uses the physical boundary constraints to correct the output of the parameter adaptive PID controller; specifically including:

[0101] Step 3.1: Construct a basic closed-loop PID control system, using the real-time temperature T at the core control measurement point upstream of the fire source. d (t) and target control temperature T t The difference is taken as the control error e(t), and the target control temperature T is set. tThe error calculation formula is:

[0102] ;

[0103] To address the discrete-time step computation requirements of FDS coupled with Python scripts, equation (5) is discretized to obtain a discrete form that can be directly used for code iteration, as shown in the following expression:

[0104] ;

[0105] In the formula, t is the current calculation time step number, u(t) is the output control law of the parameter adaptive PID controller, and K p (t), K i (t) and K d (t) represents the proportional, integral, and differential coefficients of the current step adaptive tuning;

[0106] Step 3.2: Design the differential-first module to address the overshoot problem caused by the sudden temperature rise in the early stages of a fire. The differential element is applied to the temperature feedback signal rather than the error signal to avoid drastic fluctuations in the control quantity caused by sudden changes in the setpoint. The differential-first output calculation formula is as follows:

[0107] ;

[0108] In the formula, u d (t) represents the output of the differential lead-in stage;

[0109] Step 3.3: Design an online adaptive module for all-time fire parameters. Based on fuzzy inference rules, the module uses the control error e(t) and the error change rate Δe(t) = e(t) - e(t-1) as inputs, and the adaptive adjustment amount ΔK of the PID parameters as inputs. p ΔK i ΔK d For the output, the PID parameters are tuned online in real time, taking into account the prior parameter range in step 2.

[0110] Step 3.4: Design an integral anti-saturation module, which combines integral limiting and back-calculation compensation. When the control output reaches the upper and lower limits of the air volume constraint, the accumulation of the integral term is stopped. At the same time, the cumulative error of integral saturation is eliminated by back calculation to avoid system oscillation and overshoot caused by excessive accumulation of the integral term.

[0111] Furthermore, set the target control temperature T. t If the temperature is 100℃, then the specific method for real-time online tuning of PID parameters is as follows:

[0112] (1) When |e(t)|>80℃, increase K p Decrease K iIt can quickly suppress temperature deviations and avoid over-control.

[0113] (2) When 30℃ < |e(t)| ≤ 80℃, the balance adjustment K p K i K d It balances adjustment speed and stability;

[0114] (3) When |e(t)|≤30℃, decrease K p Increase K i This improves the accuracy of steady-state control and suppresses fluctuations.

[0115] Step 4: Construct a preprocessing and anti-interference compensation module for the measurement point data, used to denoise and process outliers in the real-time monitoring data of the temperature sensor, providing a stable feedback signal for the parameter adaptive PID controller; including the following steps:

[0116] Step 4.1: Denoise the temperature data collected in real time by the temperature sensor using a first-order exponential moving average method. The calculation formula is as follows:

[0117] ;

[0118] In the formula, Tˊ d (t) represents the denoised temperature value, and α is the smoothing coefficient, with a value ranging from 0.2 to 0.5;

[0119] Step 4.2: Identify and process outliers in the denoised temperature data. Determine outliers based on the 3σ criterion. For outliers that exceed the physically reasonable range, replace them with the average of the previous and next 3 frames of data. For more than 3 consecutive frames of outlier data, automatically switch to the temperature data of the backup measuring point to achieve fault tolerance of the temperature sensor.

[0120] Step 4.3: Design an anti-interference compensation module for sudden changes in operating conditions. Based on the temperature gradient change rate upstream and downstream of the fire source, identify sudden changes in operating conditions such as a sudden increase in fire source power and flue gas backflow. When a sudden change in operating conditions is detected, automatically trigger the rapid adjustment mechanism of PID parameters, increase the weight of the proportional coefficient, shorten the system response time, and improve the anti-interference capability of the parameter adaptive PID controller.

[0121] Step 5: Construct a hybrid control objective function that integrates control error loss, physical prior constraints, and airflow smoothing constraints, and based on this hybrid control objective function, complete the parameter tuning and multi-condition simulation verification of the parameter adaptive PID controller; specifically including:

[0122] Step 5.1: Construct the control error loss term, and use the sum of squared errors index to measure control accuracy. The calculation formula is as follows:

[0123] ;

[0124] In the formula, N is the total simulation time steps;

[0125] Step 5.2: Construct a physical prior constraint loss term to penalize control quantities that exceed the physical prior boundary. The calculation formula is as follows:

[0126] ;

[0127] In the formula, V(t) is the volumetric flow rate control value of the fan at time t, and λ phy This is the physical constraint weighting coefficient, with a value ranging from 0.5 to 1.0;

[0128] Step 5.3: Construct a smoothing loss term for airflow to suppress drastic fluctuations in the fan's volumetric flow rate and avoid frequent start-ups and shutdowns and equipment damage. The calculation formula is as follows:

[0129] ;

[0130] In the formula, λ smooth The smoothing constraint weight coefficients are set to a value between 0.001 and 0.01.

[0131] Step 5.4: The mixed control objective function is a weighted sum of the loss terms, calculated using the following formula:

[0132] ;

[0133] Step 5.5: Based on multi-condition simulation data, with the minimization of the hybrid control objective function as the optimization objective, complete the initial tuning of the PID parameters, determine the initial values ​​of the parameters at each stage and the membership function of the fuzzy inference rule; through batch simulations of various fire source locations, various fire source powers and various wind speed conditions, complete the generalization verification and parameter optimization of the parameter adaptive PID controller.

[0134] Step 6: Based on the tuned parameter adaptive PID controller, real-time closed-loop intelligent control of the volumetric flow rate of the tunnel fire exhaust fan is achieved through the Python-FDS bidirectional coupling interface, specifically including:

[0135] Step 6.1: Load the tuned parameters of the adaptive PID controller, physical prior constraint parameters, data preprocessing smoothing coefficients, and standardized parameters;

[0136] Step 6.2: Extract real-time monitoring data from the temperature sensor in the tunnel step by step through the Python-FDS bidirectional coupling interface. After noise reduction preprocessing and outlier processing, input the data into the parameter adaptive PID controller.

[0137] Step 6.3: The parameter adaptive PID controller completes the online adaptive tuning of PID parameters based on the control error and error change rate at the current moment, calculates the output control rate, and converts it into the target volume flow rate of the fan at the next moment after correction by physical prior boundary constraints.

[0138] Step 6.4: Update the target volumetric flow rate to the VENT boundary of the FDS numerical model for tunnel fire in real time through the Python-FDS bidirectional coupling interface, complete the CFD calculation and control iteration for the next time step, and realize the real-time closed-loop control of the fan volumetric flow rate.

[0139] Step 6.5: Real-time storage of temperature timing data at measurement points, fan volumetric flow control commands, and PID parameter adaptive adjustment data during the control process; generation of control effect analysis reports and visualization curves; and simultaneous output to the tunnel fire emergency prevention and control platform. (See [link to relevant documentation]). Figure 4 .

[0140] 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 method for intelligent control of tunnel fire smoke exhaust fans based on physical priors and adaptive PID, characterized in that, Includes the following steps: Step 1: Construct a tunnel fire FDS numerical model with dynamic ventilation boundaries, and build a Python-FDS bidirectional coupling interface between the tunnel fire FDS numerical model and the control algorithm to achieve synchronous iteration of CFD numerical calculation and control algorithm. Step 2: Construct a physical prior system for smoke exhaust control in tunnel fires, generate physical boundary constraints to constrain the output of the fan control, and tuning prior rules to limit the parameters of the adaptive PID controller. Step 3: Design a parameter adaptive PID controller. The parameter adaptive PID controller integrates a derivative-first module, a fire-fighting all-time parameter online adaptive module, and an integral anti-saturation module, and uses the physical boundary constraints to correct the output of the parameter adaptive PID controller. Step 4: Construct a preprocessing and anti-interference compensation module for the measurement point data, which is used to denoise and process outliers in the real-time monitoring data of the temperature sensor, and provide a stable feedback signal for the parameter adaptive PID controller; Step 5: Construct a hybrid control objective function that integrates control error loss, physical prior constraints, and air volume smoothing constraints, and complete the parameter tuning and multi-condition simulation verification of the parameter adaptive PID controller based on this hybrid control objective function; Step 6: Based on the tuned parameter adaptive PID controller, the real-time closed-loop intelligent control of the volumetric flow rate of the tunnel fire smoke exhaust fan is realized through the Python-FDS bidirectional coupling interface.

2. The intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID as described in claim 1, characterized in that, Step 1, the construction of the FDS numerical model for tunnel fires and the establishment of the Python-FDS bidirectional coupling interface specifically include: Step 1.1: Establish a tunnel fire FDS numerical model that matches the actual engineering situation, and set up tunnel fire simulation scenarios with multiple fire source locations, multiple fire source power levels, and multiple longitudinal ventilation wind speed levels; Step 1.2: Adopt a dynamic grid division strategy, using the first grid size in the area near the fire source and fan, and the standard grid in the remaining areas; the first grid size is smaller than the standard grid size; Step 1.3: Configure the fan and define its entire air supply section as the VENT boundary, set its initial volumetric flow rate to the initial shutdown state, and set upper and lower limits for the fan's volumetric flow rate as hard constraints. Step 1.4: Install multiple temperature sensors at fixed intervals along the longitudinal direction of the tunnel, and set up the main feedback measuring point on the ceiling upstream of the fire source, while setting up multiple backup measuring points upstream and downstream of the fire source. Step 1.5: Synchronously execute the following Python script at each calculation time step: Extract real-time temperature data from the temperature sensor from the FDS numerical model of the tunnel fire, input it into the parameter adaptive PID controller, and update the target fan volume flow rate output by the parameter adaptive PID controller to the flow parameters at the VENT boundary in real time, so as to achieve continuous control that is completely synchronized with the FDS calculation step.

3. The intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID as described in claim 1, characterized in that, In step 2, the physical prior system includes the prior length of flue gas propagation, the physical boundary constraints for airflow control, and the tuning prior rules; wherein, The physical prior for the smoke propagation length is calculated by combining the critical temperature rise threshold method and the thermal center of mass method; the generated prior length L of the smoke propagation upstream of the fire source is... phy (t), the calculation formula is: ； In the formula, L coarse (t) represents the coarse location length of the flue gas front calculated using the critical temperature rise threshold method, L cm (t) represents the flue gas propagation length calculated using the thermal center-of-mass method, w coarse For the weight of the thick front, w cm The weight of the thermal centroid is w. coarse +w cm =1; The physical boundary constraints for airflow control generate dynamic upper and lower limits for fan volumetric flow rate based on the prior length of smoke propagation, fire source power, and tunnel ventilation characteristics. The calculation formula is as follows: ； ； In the formula, V rated Q(t) represents the rated maximum volumetric flow rate of the fan; Q(t) represents the power of the fire source at time t; k1 and k2 are the tunnel ventilation characteristic coefficients, which are determined by multi-condition simulation fitting. The aforementioned prior rules are based on the full-time evolution characteristics of tunnel fires, dividing the fire process into an initial growth period, a dynamic development period, and a quasi-steady-state control period, and setting prior intervals for the proportional coefficient, integral coefficient, and differential coefficient of each stage.

4. The intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID as described in claim 3, characterized in that, Step 3, designing the parameter adaptive PID controller specifically includes: Step 3.1: Construct a basic closed-loop PID control system, using the real-time temperature T at the core control measurement point upstream of the fire source. d (t) and target control temperature T t The difference is taken as the control error e(t), and the target control temperature T is set. t The error calculation formula is: ； To address the discrete-time step computation requirements of FDS coupled with Python scripts, the above equation is discretized to obtain a discrete form that can be directly used for code iteration, as shown in the following expression: ； In the formula, t is the current calculation time step number, u(t) is the output control law of the parameter adaptive PID controller, and K p (t), K i (t) and K d (t) represents the proportional, integral, and differential coefficients of the current step adaptive tuning; Step 3.2: Design the differential-first module to address the overshoot problem caused by the sudden temperature rise in the early stages of a fire. The differential element is applied to the temperature feedback signal rather than the error signal to avoid drastic fluctuations in the control quantity caused by sudden changes in the setpoint. The differential-first output calculation formula is as follows: ； In the formula, u d (t) represents the output of the differential lead-in stage; Step 3.3: Design an online adaptive module for all-time fire parameters. Based on fuzzy inference rules, the module uses the control error e(t) and the error change rate Δe(t) = e(t) - e(t-1) as inputs, and the adaptive adjustment amount ΔK of the PID parameters as inputs. p ΔK i ΔK d For the output, the PID parameters are tuned online in real time, taking into account the prior parameter range in step 2. Step 3.4: Design an integral anti-saturation module, which combines integral limiting and back-calculation compensation. When the control output reaches the upper and lower limits of the air volume constraint, the accumulation of the integral term is stopped. At the same time, the cumulative error of integral saturation is eliminated by back calculation to avoid system oscillation and overshoot caused by excessive accumulation of the integral term.

5. The intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID as described in claim 4, characterized in that, Set target control temperature T t If the temperature is 100℃, then the specific method for real-time online tuning of PID parameters is as follows: (1) When |e(t)|>80℃, increase K p Decrease K i It can quickly suppress temperature deviations and avoid over-control. (2) When 30℃ < |e(t)| ≤ 80℃, the balance adjustment K p K i K d It balances adjustment speed and stability; (3) When |e(t)|≤30℃, decrease K p Increase K i This improves the accuracy of steady-state control and suppresses fluctuations.

6. The intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID as described in claim 1, characterized in that, Step 4, the construction of the preprocessing and anti-interference compensation module for the measurement point data includes the following steps: Step 4.1: Denoise the temperature data collected in real time by the temperature sensor using a first-order exponential moving average method. The calculation formula is as follows: ； In the formula, Tˊ d (t) represents the denoised temperature value, and α is the smoothing coefficient, with a value ranging from 0.2 to 0.5; Step 4.2: Identify and process outliers in the denoised temperature data. Determine outliers based on the 3σ criterion. For outliers that exceed the physically reasonable range, replace them with the average of the previous and next 3 frames of data. For more than 3 consecutive frames of outlier data, automatically switch to the temperature data of the backup measuring point to achieve fault tolerance of the temperature sensor. Step 4.3: Design an anti-interference compensation module for sudden changes in operating conditions. Based on the temperature gradient change rate upstream and downstream of the fire source, identify sudden changes in operating conditions such as a sudden increase in fire source power and flue gas backflow. When a sudden change in operating conditions is detected, automatically trigger the rapid adjustment mechanism of PID parameters, increase the weight of the proportional coefficient, shorten the system response time, and improve the anti-interference capability of the parameter adaptive PID controller.

7. The intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID as described in claim 1, characterized in that, Step 5, the construction and parameter tuning of the hybrid control objective function specifically includes: Step 5.1: Construct the control error loss term, and use the sum of squared errors index to measure control accuracy. The calculation formula is as follows: ； In the formula, N is the total simulation time steps; Step 5.2: Construct a physical prior constraint loss term to penalize control quantities that exceed the physical prior boundary. The calculation formula is as follows: ； In the formula, V(t) is the volumetric flow rate control value of the fan at time t, and λ phy This is the physical constraint weighting coefficient, with a value ranging from 0.5 to 1.0; Step 5.3: Construct a smoothing loss term for airflow to suppress drastic fluctuations in the fan's volumetric flow rate and avoid frequent start-ups and shutdowns and equipment damage. The calculation formula is as follows: ； In the formula, λ smooth The smoothing constraint weight coefficients are set to a value between 0.001 and 0.

01. Step 5.4: The mixed control objective function is a weighted sum of the loss terms, calculated using the following formula: ； Step 5.5: Based on multi-condition simulation data, with the minimization of the hybrid control objective function as the optimization objective, complete the initial tuning of the PID parameters, determine the initial values ​​of the parameters at each stage and the membership function of the fuzzy inference rule; through batch simulations of various fire source locations, various fire source powers and various wind speed conditions, complete the generalization verification and parameter optimization of the parameter adaptive PID controller.

8. The intelligent control method for tunnel fire smoke exhaust fans based on physical priors and adaptive PID as described in claim 2, characterized in that, Step 6, the real-time closed-loop intelligent control specifically includes: Step 6.1: Load the tuned parameters of the adaptive PID controller, physical prior constraint parameters, data preprocessing smoothing coefficients, and standardized parameters; Step 6.2: Extract real-time monitoring data from the temperature sensor in the tunnel step by step through the Python-FDS bidirectional coupling interface. After noise reduction preprocessing and outlier processing, input the data into the parameter adaptive PID controller. Step 6.3: The parameter adaptive PID controller completes the online adaptive tuning of PID parameters based on the control error and error change rate at the current moment, calculates the output control rate, and converts it into the target volume flow rate of the fan at the next moment after correction by physical prior boundary constraints. Step 6.4: Update the target volumetric flow rate to the VENT boundary of the FDS numerical model for tunnel fire in real time through the Python-FDS bidirectional coupling interface, complete the CFD calculation and control iteration for the next time step, and realize the real-time closed-loop control of the fan volumetric flow rate. Step 6.5: Real-time storage of temperature timing data at measurement points, fan volume flow control commands, and PID parameter adaptive adjustment data during the control process; generation of control effect analysis reports and visualization curves; and synchronous output to the tunnel fire emergency prevention and control platform.

Images