A system, method, apparatus and storage medium for calibrating an aspirating smoke detector

Abstract

The application discloses a kind of aspirating smoke detector correction systems, including data acquisition layer, digital twin modeling layer, pipe network parameterization and optimization layer and virtual-real linkage correction layer, data flow is respectively connected the data acquisition layer, digital twin modeling layer, pipe network parameterization and optimization layer and virtual-real linkage correction layer, data acquisition layer is configured laser radar equipment and multidimensional sensor network, digital twin modeling layer is communicated with data acquisition layer, for building and physical protection space 1:1 high-precision mapping digital twin model;Pipe network parameterization and optimization layer are used for realizing sampling pipe network parameterization modeling, intelligent iterative optimization;Virtual-real linkage correction layer is used for realizing the two-way linkage of virtual model and physical pipe network.The aspirating smoke detector correction method of the application has strong low-temperature adaptability, completely solves the pain point of extreme working condition, has high design precision, and greatly improves detection performance;Full life cycle intelligent operation and maintenance, cost reduction and efficiency improvement are remarkable.

Description

Technical Field

[0001] This invention relates to the field of fire protection, and more particularly to a calibration system, method, and storage medium for an aspirating smoke detector. Background Technology

[0002] In the field of fire protection, aspirating smoke detectors (ASDs) actively collect air samples through sampling pipelines. With their advantages of high sensitivity, long sampling distance, and strong anti-interference ability, they are widely used in various high-safety-risk and special working conditions fire detection scenarios.

[0003] Chinese Patent No. 2005100469142 provides a method for design verification and optimization of sampling pipeline network for aspirating smoke fire detection system. This method does not consider dynamic working conditions such as airflow disturbance in complex spaces and intermittent fluctuations of low-temperature airflow, and it does not consider the matching degree and toughness of pipeline network design under various extreme working conditions such as material stress.

[0004] The existing design and maintenance technology for sampling pipeline networks of aspirating smoke detectors has significant technical shortcomings in extreme low-temperature environments such as cold storage, as follows: 1. Design accuracy is highly dependent on human experience. Traditional sampling pipeline design often uses fixed empirical formulas for parameter calculation, which cannot accurately simulate dynamic conditions such as airflow disturbances and intermittent fluctuations of low-temperature airflow in complex spaces. This can easily lead to blind spots in fire detection. At the same time, existing technologies have not conducted targeted material mechanics and thermodynamic matching calculations for extreme low-temperature scenarios of -40℃, resulting in extremely poor adaptability of pipeline design. 2. Extremely poor environmental adaptability. In practical applications, the frequent start-up and shutdown of cold storage air conditioners and refrigeration units will cause dynamic fluctuations in airflow, temperature and humidity. Traditional static pipe network design schemes cannot adjust parameters in real time to keep up with environmental changes. Moreover, conventional PVC pipes are prone to low-temperature brittle cracking under low-temperature conditions. When low-temperature cold air is drawn into the detector, it is very easy to cause condensation and frost inside the equipment, resulting in detector failure and detection failure. 3. Disconnect between virtual and real systems leads to high trial-and-error costs in engineering. Existing pipeline optimization schemes mostly use offline simulation software for simulation design, which only stays at the virtual simulation level and lacks a closed-loop feedback mechanism with physical hardware and actual operating conditions. This results in a serious disconnect between the simulation model and the actual on-site operating conditions, poor implementation of the design scheme, and extremely high costs for repeated debugging and trial and error.

[0005] Therefore, it is necessary to develop a calibration method for aspirating smoke detectors to overcome the shortcomings of existing technologies, such as low design accuracy of sampling pipeline networks in low-temperature environments, poor environmental adaptability, disconnect between virtual and real models, and high operation and maintenance costs. Summary of the Invention

[0006] In a first aspect, the present invention provides a calibration system for an inhalation-type smoke detector, comprising an inhalation-type smoke detector, further comprising a data acquisition layer, a digital twin modeling layer, a pipeline parameterization and optimization layer, and a virtual-real linkage calibration layer, and further comprising a data stream, wherein the data stream is respectively connected to the data acquisition layer, the digital twin modeling layer, the pipeline parameterization and optimization layer, and the virtual-real linkage calibration layer; the data acquisition layer is configured with a lidar device and a multi-dimensional sensor network for acquiring three-dimensional geometric structure data of the target protection area, real-time environmental airflow parameters, temperature and humidity parameters, and pipeline operation status data; the digital twin modeling layer is communicatively connected to the data acquisition layer for constructing a digital twin model that is 1:1 highly accurate mapped to the physical protection space; the pipeline parameterization and optimization layer is used to realize parameterized modeling of the sampling pipeline network, intelligent iterative optimization, and define core design parameters and low-temperature adaptation parameters of the pipeline network; the virtual-real linkage calibration layer is configured with an electric regulating valve, a variable aperture sampler, a low-temperature air heater, and a pipeline expansion and contraction compensation device to realize bidirectional linkage between the virtual model and the physical pipeline network.

[0007] Furthermore, the digital twin modeling layer further includes a geometric modeling module, a physical attribute mapping module, and a dynamic behavior modeling module.

[0008] Furthermore, the pipeline parameterization and optimization layer further includes a parameterization engine, a machine learning agent model unit, and an intelligent optimization module.

[0009] Secondly, the present invention provides a calibration method for an inhalation-type smoke detector, comprising the following steps: Step 201: Multi-source data fusion and digital twin modeling, specifically: collect geometric data of the three-dimensional structure, insulation layer and equipment layout of the target cold storage building through lidar, and collect multi-source data in real time through sensor network; fuse multi-source data to construct a 1:1 digital twin model including geometric structure, physical properties and low-temperature dynamic behavior, embed a low-temperature airflow intermittent change model and a condensation and icing prediction model, and complete the initialization of model boundary conditions; Step 202: Parametric modeling and initial scheme generation, specifically: calculate the friction loss along the pipeline based on the fluid mechanics continuity equation and Darcy-Weisbach formula, and calculate the overall resistance of the pipeline based on the local resistance coefficient method; calculate the thermal expansion of the ABS sampling pipeline, set pipeline expansion compensation, pressure drop balance, and flow balance constraints, and generate an initial design scheme for the pipeline adapted to the low temperature environment; Step 203: Rapid multi-objective optimization driven by the surrogate model, specifically: calling the pre-trained machine learning surrogate model to quickly predict the detector response time, smoke capture efficiency and pipeline icing risk corresponding to different pipeline parameters, setting multi-objective optimization constraints, combining genetic algorithm for iterative optimization, and outputting the globally optimal pipeline aperture, aperture position, pipe diameter and expansion joint layout scheme.

[0010] Step 204: Precise alignment and physical deployment of virtual and real systems, specifically: Based on the optimal design scheme, complete the on-site installation and deployment of physical sampling pipe network, heater, expansion compensation device, and electric regulating valve; collect the actual physical parameters of pipe material roughness, pipe elbow resistance, opening burr error and equipment installation deviation on site; synchronously update the boundary conditions and parameters of the digital twin model; eliminate the initial deviation of the virtual and real models; and achieve high-precision alignment of virtual and real working conditions. Step 205: Dynamic correction and closed-loop operation and maintenance throughout the entire life cycle. Specifically, this involves: real-time collection of pipeline operation data, environmental dynamic data, and detector alarm data; continuous comparison of the deviation between the theoretical prediction data of the digital twin model and the actual operation data of the physical equipment; when the deviation exceeds the preset threshold, the system determines that there is dust accumulation, blockage, or sudden change in environmental conditions in the pipeline, automatically corrects the core parameters of the friction coefficient of the virtual model, and adaptively adjusts the opening of the electric regulating valve and the power of the air heater to compensate for the airflow loss in the pipeline and avoid the risk of low-temperature freezing and condensation, thereby realizing adaptive optimization closed-loop management of the pipeline throughout its entire life cycle.

[0011] Furthermore, the straight-line distance between the heater and the inhalation-type smoke detector is 250~350mm.

[0012] Thirdly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the inhalation-type smoke detector calibration method as described in any of the second aspects.

[0013] Fourthly, the present invention provides a calibration device for an inhalation-type smoke detector, comprising: One or more processors; Memory; and One or more computer programs, wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, characterized in that, when the processor executes the computer program, it implements the steps of the inhalation smoke detector calibration method as described in any of the second aspects.

[0014] This invention, by introducing a machine learning proxy model with physical calibration features and a virtual-real linkage closed-loop correction mechanism, combined with dedicated low-temperature fire protection hardware linkage control logic, completely solves technical problems such as pipeline brittleness, detector condensation, detection blind zones, and poor dynamic adaptability in extreme low-temperature scenarios. It achieves intelligent adaptive design and precise operation and maintenance management throughout the entire lifecycle of the sampling pipeline network. It boasts strong low-temperature adaptability, completely resolving pain points in extreme operating conditions: high design accuracy and significantly improved detection performance; intelligent operation and maintenance throughout the entire lifecycle, resulting in significant cost reduction and efficiency improvement. Attached Figure Description

[0015] Figure 1 : Architecture diagram of the inhalation-type smoke detector calibration system of this invention.

[0016] Figure 2 Flowchart of the inhalation-type smoke detector calibration method of the present invention.

[0017] Figure 3 : Schematic diagram of the deviation correction principle of the inhalation type smoke detector correction method of the present invention.

[0018] Figure 4 : A schematic diagram of the composition of the inhalation-type smoke detector calibration device of the present invention. Detailed Implementation

[0019] To make the technical problems, technical solutions and beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] refer to Figure 1 The architecture diagram of the inhalation-type smoke detector calibration system of this invention includes a four-layer architecture: a data acquisition layer 101, a digital twin modeling layer 102, a pipeline parameterization and optimization layer 103, and a virtual-real linkage calibration layer 104. The data flow 105 operates in a closed-loop linkage and collaborative manner at each of the above layers, wherein: The data acquisition layer 101 is equipped with a LiDAR device and a multi-dimensional sensor network to acquire three-dimensional geometric structure data of the target protection area, real-time environmental airflow parameters, temperature and humidity parameters, and pipeline operation status data. For extreme low-temperature scenarios in cold storage, the sensor network is additionally equipped with a thin-film condensation sensor, a fiber optic grating icing sensor, and a refrigeration device parameter acquisition module. It can collect core operating condition data in real time, such as the distribution of the cold storage insulation layer, the start-up and shutdown parameters of the refrigeration device, the surface temperature of the pipeline, the condensation and icing phase change state, and the indoor and outdoor temperature difference, providing accurate data support for model building and parameter optimization. The digital twin modeling layer 102 is communicatively connected to the data acquisition layer and is used to construct a digital twin model that is 1:1 highly accurate and mapped to the physical protection space. The digital twin modeling layer 102 further includes a geometric modeling module, a physical attribute mapping module, and a dynamic behavior modeling module (not shown in the figure). The core function of the digital twin modeling layer 102 is to embed a high-precision CFD simulation kernel and an FDS fire dynamics simulation engine. For the special low-temperature working conditions of cold storage, it embeds an unsteady heat and moisture transfer model, a low-temperature airflow intermittent change model, and a pipeline condensation and icing prediction model. It can accurately simulate the airflow fluctuation, pipeline thermal deformation, and condensation and frost evolution process under low-temperature environment, and predict the risk of pipeline icing and detector condensation in advance. The pipeline parameterization and optimization layer 103 is used to realize parameterized modeling and intelligent iterative optimization of the sampled pipeline network, and to define the core design parameters and low-temperature adaptation parameters of the pipeline network. The core design parameters include pipe diameter, number of openings, opening size, thickness of pipe insulation layer, installation spacing of expansion joints, etc. The pipeline parameterization and optimization layer 103 further includes a parameterization engine, machine learning agent model unit and intelligent optimization module (not shown in the figure).

[0021] The parameterization engine accurately calculates the friction loss along the pipeline network based on the fluid mechanics continuity equation and the Darcy-Weisbach formula, and calculates the overall pipeline network resistance using the local resistance coefficient method. An air viscosity-temperature correction coefficient is introduced under low-temperature conditions to adapt to the characteristics of extreme low-temperature fluids. The machine learning proxy model unit adopts a CNN / RNN deep neural network structure. By learning from massive amounts of offline CFD simulation data, it establishes a nonlinear mapping relationship between pipeline design parameters, environmental parameters, detector response performance, and pipeline operating status, compressing the traditional CFD simulation prediction time of several hours to the millisecond level and solving the problem of lag in real-time iterative optimization. The intelligent optimization module uses a genetic algorithm with multiple optimization objectives, including minimum detector response time, minimum detection blind zone, and minimum pipeline icing risk, to achieve global optimal search for low-temperature pipeline network parameters. The virtual-real linkage correction layer 104 is equipped with actuators such as electric regulating valves, variable orifice samplers, low-temperature air heaters, and pipeline expansion and contraction compensation devices to achieve bidirectional linkage between the virtual model and the physical pipeline network. The virtual-real linkage correction layer 104 can implement the optimal pipeline network design scheme output by the virtual layer into the physical world, and at the same time collect physical feedback information such as actual installation errors, pipe material losses, operating conditions, and detector alarm data of the physical pipeline network to complete the recalibration of digital twin model parameters. When the deviation between virtual and real operating data and alarm data exceeds the threshold, the model parameters are automatically corrected, and the physical actuators are driven to adaptively adjust the pipeline network operating parameters to achieve precise synchronization of virtual and real closed loop.

[0022] refer to Figure 2 The flowchart of the inhalation-type smoke detector calibration method of the present invention includes the following steps: Step 201: Multi-source data fusion and digital twin modeling, specifically: acquiring geometric data of the target cold storage building's three-dimensional structure, insulation layer, and equipment layout using lidar; acquiring real-time operating condition data such as ambient temperature and humidity, indoor and outdoor temperature difference, refrigeration unit start-up and shutdown cycles, and pipe surface conditions using a sensor network; fusing multi-source data to construct a 1:1 digital twin model including geometric structure, physical properties, and low-temperature dynamic behavior; embedding a low-temperature airflow intermittent change model and a condensation and icing prediction model; and completing the initialization of model boundary conditions. Step 202: Parametric modeling and initial scheme generation, specifically: calculating the friction loss along the pipeline based on the fluid mechanics continuity equation and Darcy-Weisbach formula, and verifying the overall resistance of the pipeline based on the local resistance coefficient method; for the extreme low temperature condition of -40℃, introducing an air viscosity correction coefficient to correct the low temperature air fluid characteristic parameters; verifying the thermal expansion of the ABS sampling pipeline, setting pipeline expansion and contraction compensation, pressure drop balance, and flow balance constraints, and generating an initial design scheme for the pipeline adapted to the low temperature environment; Step 203: Rapid multi-objective optimization driven by surrogate model, specifically: calling a pre-trained machine learning surrogate model to replace traditional CFD simulation, quickly predicting detector response time, smoke capture efficiency and pipeline icing risk corresponding to different pipeline parameters, setting multi-objective optimization constraints, including response time ≤10s, detection blind zone ≤5%, and no pipeline icing and condensation risk, etc., combined with genetic algorithm iterative optimization, outputting the globally optimal pipeline aperture, aperture position, pipe diameter, and expansion joint layout scheme.

[0023] Step 204: Precise alignment and physical deployment of virtual and real systems, specifically: Based on the optimal design scheme, complete the on-site installation and deployment of physical sampling pipe network, heater, expansion compensation device, and electric regulating valve; collect actual physical parameters such as pipe roughness, pipe elbow resistance, opening burr error, and equipment installation deviation; synchronously update the boundary conditions and parameters of the digital twin model; eliminate the initial deviation between virtual and real models; and achieve high-precision alignment of virtual and real working conditions. Step 205: Dynamic calibration and closed-loop operation and maintenance throughout the entire lifecycle, specifically: the system collects pipeline operation data, environmental dynamic data, and detector alarm data in real time, and continuously compares the deviation between the theoretical prediction data of the digital twin model and the actual operation data of the physical equipment; further referencing Figure 3 The deviation correction principle diagram of the inhalation type smoke detector correction method of this invention shows that when the deviation exceeds the preset threshold, the system determines that there is dust accumulation, blockage or sudden change in environmental conditions in the pipeline network, automatically corrects the core parameters such as the friction coefficient of the virtual model, and adaptively adjusts the opening of the electric regulating valve and the power of the air heater to compensate for the airflow loss in the pipeline network, avoid the risk of low temperature freezing and condensation, and realize adaptive optimization closed-loop management of the pipeline network throughout its entire life cycle.

[0024] As a preferred embodiment of the present invention, the design calibration is performed using the aspirating smoke detector calibration method of the present invention, taking a cold storage scenario of -40℃ as an example for illustration: Step 1: Conduct multi-source data acquisition and digital twin model construction. Use 3D LiDAR to perform a panoramic scan of the target cold storage, acquiring millimeter-level 3D point cloud geometric data of the cold storage walls, beams, columns, insulation layers, refrigeration units, and shelving equipment to complete spatial geometric modeling. Real-time data collection of cold storage operating conditions is achieved through a sensor network: the refrigeration unit operates for 15 minutes and stops for 45 minutes; the lowest internal temperature is -40℃, the outdoor temperature is 40℃, and the maximum indoor-outdoor temperature difference ΔT = 80℃. Physical property configuration: Industrial ABS antifreeze pipes are selected as the main material of the sampling pipe network. The thermal expansion coefficient of ABS pipes is locked in the physical property mapping module of the digital twin model as α=10.1×10 / ℃·m. Dynamic behavior configuration: The FDS fire dynamics engine and low temperature airflow intermittent change model are embedded. The pipe dew point warning threshold is set to -15℃, and the condensation and icing prediction logic is initialized.

[0025] Step 2: Perform parametric modeling and generate the initial scheme for the cryogenic pipeline network. First, define the basic parameters of the pipeline network: main pipe diameter 25mm, effective length of a single pipe section L=40m, and calculate the cryogenic deformation parameters based on the formula for calculating pipe expansion / contraction ΔL. ; Because the calculated expansion amount ΔL=323mm>50mm preset threshold, the system forces the installation of a set of pipeline expansion compensation devices in the middle section of the pipeline to offset the low temperature cold and heat deformation stress and prevent the pipeline from becoming brittle. Fluid resistance calculation: Adapted to -40℃ low-temperature conditions, an air viscosity correction factor μ=1.60×10 Pa·s is introduced. The friction resistance along the pipeline network is calculated based on the Darcy-Weisbach formula, and the resistance calibration of the entire pipeline section is completed using the local resistance coefficient method. The system calculation shows that the initial orifice diameter range for a total sampling flow rate of 0.5~1.5 m / min is 3.0 mm~4.5 mm. The orifice diameter difference between adjacent sampling orifices is controlled to be ≤0.5 mm, ensuring that the flow rate deviation at each orifice location is <5%, achieving uniform sampling across the entire area.

[0026] Step 3: Machine learning agent model drives multi-objective rapid optimization. The pre-trained machine learning agent model is called, and 100 sets of working condition samples covering ventilation rate of 0.5~10 times / h and fire heat release rate of 5~50kW are used as the training basis. Genetic algorithm is used to carry out multi-objective iterative optimization. For the low temperature and high freezing risk scenario of cold storage, the freezing risk weight factor is set to 0.5 to balance detection sensitivity and antifreeze requirements. The optimized solution is as follows: In areas with large airflow fluctuations and high risk of icing, such as the air outlet of the refrigeration unit, the sampling hole density is 2m / hole with a hole diameter of 4.0mm; In the central area of ​​the cold storage where the low temperature is stable and the risk is low, the sampling hole density is 4m / hole with a hole diameter of 3.5mm, which balances full detection coverage and low temperature operation stability.

[0027] Step 4: Physical hardware deployment and virtual-physical model calibration. Complete the sampling pipeline laying according to the optimized plan, and simultaneously complete the low-temperature hardware deployment: Install an air heater with a rated power of 200W at the location where the sampling pipe exits the cold storage and is close to the detector. The straight-line distance between the heater and the aspirating smoke detector is controlled at 300mm, which is in the optimal range of 250~350mm. This can stably heat the -40℃ low-temperature intake air to above 5℃ without the risk of high temperature damage to the detector circuit. On-site parameter benchmarking: Collect real parameters such as roughness increment, burr error of opening, and actual resistance of elbow generated during on-site pipeline transportation and installation, and update the boundary conditions of the digital twin model simultaneously to eliminate the initial deviation between the virtual and real models and achieve 1:1 precise alignment.

[0028] Step 5: Dynamic deviation correction and closed-loop operation and maintenance throughout the entire life cycle. The system monitors the pipeline network's operating status and environmental parameters in real time throughout the entire process, and executes dual logic for low-temperature antifreeze and fault compensation. Low temperature antifreeze warning: When the thin film condensation sensor detects that the pipe wall temperature is ≤-15℃ and the ambient relative humidity is ≥90%, the system will forcibly start the air heater to stabilize the airflow temperature in the pipe at 5~15℃, completely eliminating the fault of condensation and frost inside the detector. Adaptive correction of flow deviation: The system presets an alarm time deviation threshold of 10%. Actual test verification: The theoretical alarm time of the digital twin model is 50s, and the actual alarm time of the physical detector is 55s. The deviation just reaches the 10% threshold. The system determines that there is slight dust accumulation and airflow loss in the branch. Then it drives the electric regulating valve to increase the current opening by 20%, which is in the optimal adjustment range of 15%~25%, to compensate for the airflow loss in the pipeline and restore the detector response performance to the standard range. Continuous iterative optimization: The system feeds back the environmental conditions and operational correction data collected over a long period of time to the machine learning agent model, dynamically fine-tunes the icing risk weight factor and resistance correction parameters, and realizes adaptive evolution throughout the entire life cycle of pipeline design and operation and maintenance.

[0029] Further reference Figure 4 A schematic diagram of the composition of the inhalation smoke detector calibration device of the present invention is shown. The inhalation smoke detector calibration device 10 further includes one or more memories 20 and one or more processors 30, wherein the one or more computer programs are stored in the memories 20 and configured to be executed by the one or more processors 30. When the processors 30 execute the computer programs, they implement the steps of the inhalation smoke detector calibration method.

[0030] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0031] 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, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A calibration system for an inhalation-type smoke detector, comprising an inhalation-type smoke detector, characterized in that, The system further includes a data acquisition layer, a digital twin modeling layer, a pipeline parameterization and optimization layer, and a virtual-real linkage correction layer. It also includes a data stream, which connects to the data acquisition layer, digital twin modeling layer, pipeline parameterization and optimization layer, and virtual-real linkage correction layer. The data acquisition layer is equipped with lidar equipment and a multi-dimensional sensor network to acquire three-dimensional geometric structure data of the target protection area, real-time environmental airflow parameters, temperature and humidity parameters, and pipeline operation status data. The digital twin modeling layer is communicatively connected to the data acquisition layer to construct a digital twin model that maps to the physical protection space at a 1:1 high precision. The pipeline parameterization and optimization layer is used to achieve parameterized modeling of the sampling pipeline network, intelligent iterative optimization, and to define core design parameters and low-temperature adaptation parameters for the pipeline network. The virtual-real linkage correction layer is equipped with electric regulating valves, variable orifice samplers, low-temperature air heaters, and pipeline expansion and contraction compensation devices to achieve bidirectional linkage between the virtual model and the physical pipeline network.

2. The inhalation-type smoke detector calibration system as described in claim 1, characterized in that, The digital twin modeling layer further includes a geometric modeling module, a physical attribute mapping module, and a dynamic behavior modeling module.

3. The inhalation-type smoke detector calibration system as described in claim 1, characterized in that, The pipeline parameterization and optimization layer further includes a parameterization engine, a machine learning agent model unit, and an intelligent optimization module.

4. A calibration method for an inhalation-type smoke detector, characterized in that, Includes the following steps: Step 201: Multi-source data fusion and digital twin modeling, specifically: collect geometric data of the three-dimensional structure, insulation layer and equipment layout of the target cold storage building through lidar, and collect multi-source data in real time through sensor network; fuse multi-source data to construct a 1:1 digital twin model including geometric structure, physical properties and low-temperature dynamic behavior, embed a low-temperature airflow intermittent change model and a condensation and icing prediction model, and complete the initialization of model boundary conditions; Step 202: Parametric modeling and initial scheme generation, specifically: calculate the friction loss along the pipeline based on the fluid mechanics continuity equation and Darcy-Weisbach formula, and calculate the overall resistance of the pipeline based on the local resistance coefficient method; calculate the thermal expansion of the ABS sampling pipeline, set pipeline expansion compensation, pressure drop balance, and flow balance constraints, and generate an initial design scheme for the pipeline adapted to the low temperature environment; Step 203: Rapid multi-objective optimization driven by surrogate model, specifically: calling the pre-trained machine learning surrogate model to quickly predict the detector response time, smoke capture efficiency and pipeline icing risk corresponding to different pipeline parameters, setting multi-objective optimization constraints, combining genetic algorithm for iterative optimization, and outputting the globally optimal pipeline aperture, aperture position, pipe diameter and expansion joint layout scheme. Step 204: Precise alignment and physical deployment of virtual and real systems, specifically: Based on the optimal design scheme, complete the on-site installation and deployment of physical sampling pipe network, heater, expansion compensation device, and electric regulating valve; collect the actual physical parameters of pipe material roughness, pipe elbow resistance, opening burr error and equipment installation deviation on site; synchronously update the boundary conditions and parameters of the digital twin model; eliminate the initial deviation of the virtual and real models; and achieve high-precision alignment of virtual and real working conditions. Step 205: Dynamic correction and closed-loop operation and maintenance throughout the entire life cycle. Specifically, this involves: real-time collection of pipeline operation data, environmental dynamic data, and detector alarm data; continuous comparison of the deviation between the theoretical prediction data of the digital twin model and the actual operation data of the physical equipment; when the deviation exceeds the preset threshold, the system determines that there is dust accumulation, blockage, or sudden change in environmental conditions in the pipeline, automatically corrects the core parameters of the friction coefficient of the virtual model, and adaptively adjusts the opening of the electric regulating valve and the power of the air heater to compensate for the airflow loss in the pipeline and avoid the risk of low-temperature freezing and condensation, thereby realizing adaptive optimization closed-loop management of the pipeline throughout its entire life cycle.

5. The calibration method for an inhalation-type smoke detector as described in claim 4, characterized in that, The The straight-line distance between the heater and the inhalation-type smoke detector is 250~350mm.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the inhalation-type smoke detector calibration method as described in any one of claims 4 to 5.

7. A calibration device for an inhalation-type smoke detector, comprising: One or more processors; Memory; as well as One or more computer programs, wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, characterized in that, when the processor executes the computer program, it implements the steps of the inhalation smoke detector calibration method as described in any one of claims 4 to 5.

Images