Gas sensor layout method

Abstract

The invention discloses a method for arranging gas sensors. The method comprises the following steps: collecting related gas parameters of a subway tunnel during operation; performing risk evaluation on each point location by using a probability-influence-based risk evaluation method, and determining an importance index of each node; according to the subway tunnel ventilation network diagram and the determined monitoring effective level, a subway tunnel gas monitoring system sensor layout optimization model based on quantitative risk evaluation is generated through a maximum coverage site selection model; the optimization model is solved based on a branch pricing algorithm, an optimal gas sensor layout strategy is obtained, and the gas sensor layout strategy in the subway tunnel comprises determination of gas sensor layout points. And a distributed sensor arrangement strategy is adopted, so that the coverage capability of the system on each part of the subway tunnel is enhanced, and the flexibility and expansibility of the whole system are improved. The safety of the subway tunnel is improved, and the economic cost of the monitoring system is reduced.

Description

Technical Field

[0001] The present invention relates to a method for testing or analyzing a material by means of measuring the chemical or physical properties of the material, and its IPC classification number is G01N, and in particular, it relates to a method for arranging a gas sensor. Background Art

[0002] Subway tunnels are generally equipped with advanced safety monitoring systems. The main purpose of these systems is to monitor the basic safety parameters and status of subway tunnels in real time and dynamically during construction and operation. Gas detection devices are one of the most important links in the safety monitoring system of subway tunnels. The number and layout of sensors are crucial to preventing accidents: too few sensors cannot effectively prevent accidents, while too many sensors will lead to excessive investment in the monitoring system. In fact, the function and efficiency of subway tunnel safety monitoring systems depend to a large extent on the precise location and layout of sensors. Even with the same number of sensors, different layout strategies will produce completely different monitoring effects. Despite this, there are relatively few research reports on the optimal layout of gas sensors in subway ventilation tunnels at home and abroad, especially on the subway operation stage.

[0003] Zhao Dan et al. studied a method for deploying gas sensors (patent publication number: CN108918813A). This method first constructs a complete mine ventilation network directed graph containing virtual nodes based on the mine ventilation network diagram and the relevant specifications and standards of the gas sensor. Subsequently, the relevant algorithm is used to generate the monitoring time matrix of the node. Next, the shortest alarm time of the gas sensor is set, and an effective monitoring matrix of the node is established. In this process, the effective monitoring matrix is ​​simplified, and the columns related to the virtual nodes and the rows they cover are removed to obtain a simplified node effective monitoring matrix. Finally, based on this simplified matrix, the objective function is determined, and the construction of the gas sensor deployment model is completed.

[0004] Although there is some research on the layout of gas sensors in mines, there is a lack of necessary research on the layout of gas sensors in subway tunnels. In view of the serious consequences that may be caused by gas leakage, it is particularly necessary to scientifically and rationally optimize the layout of gas sensors in subway tunnels. This will not only help improve the safety of subway tunnels, but also ensure the economic benefits of the monitoring system and provide solid safety guarantees for subway operations. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art. The present invention provides a distributed subway tunnel gas monitoring system sensor arrangement method to solve the problem that the existing subway tunnel gas monitoring system sensor arrangement effect is poor.

[0006] The present invention provides a method for deploying a gas sensor, comprising the following steps:

[0007] S1: Collect relevant gas parameters of subway tunnels during operation;

[0008] The gas-related data include structural parameters, operating parameters and ventilation system parameters of the subway tunnel;

[0009] S2: Use the probability-impact based risk assessment method to evaluate the risk of each point and determine the importance index of each node;

[0010] S3: Based on the subway tunnel ventilation network diagram and the determined monitoring effectiveness level, the subway tunnel gas monitoring system sensor layout optimization model based on quantitative risk assessment is generated through the maximum coverage site selection model;

[0011] In step S3, the sensor layout of the subway tunnel gas monitoring system aims to maximize the disaster preparedness target, and the objective function is:

[0012]

[0013] Where: i is the importance index of each demand node i obtained based on the probability-impact risk assessment model. The objective function is to maximize the disaster preparedness target of the entire region. i Indicates whether the gas situation at demand point i can be effectively monitored. If the monitored y i =1, otherwise y i =0.

[0014] The constraints of step S3 are as follows:

[0015] satisfy:

[0016]

[0017] Where: Formula (2) represents the selection of P from the candidate monitoring point set J j Gas monitoring points, x j Indicates whether the candidate monitoring point j is selected. If x is selected j =1, otherwise x j =0,y i Indicates whether the gas situation at demand point i can be effectively monitored. If the monitored y i =1, otherwise y i =0; In this model, it is assumed that the cost of installing gas sensors at each alternative layout point is the same. ij is an element in the node effective monitoring matrix or branch effective monitoring matrix at a given monitoring effectiveness level, cij ∈{0,1}, let c ij = 1, it is assumed that the sensor at node i can monitor the gas situation at node j in a timely and effective manner; otherwise, let c ij = 0, it is considered that the sensor at node i cannot monitor the gas situation at node j, and the effective monitoring matrix C = [c ij ] is constructed based on the node monitoring time matrix T.

[0018] S4: Solving the optimization model described in step S3 based on the branch pricing algorithm to obtain the optimal gas sensor layout strategy, wherein the gas sensor layout strategy in the subway tunnel includes determining the gas sensor deployment points.

[0019] The tunnel structural parameters include the length, width, height, slope, cross-sectional shape, entrance and exit locations, and layout of key electrical equipment of the subway tunnel;

[0020] The operation parameters are information such as the subway operation schedule, vehicle operation frequency, and passenger flow;

[0021] The ventilation system parameters are to obtain data such as the ventilation mode, ventilation resistance, wind speed, wind direction, etc. of the tunnel.

[0022] Furthermore, in step S1, the optimization of the sensor layout of the gas monitoring system in the subway tunnel based on quantitative risk assessment is to conduct risk assessment on each point using a probability (Probability)-impact (Impart) risk assessment method under the condition that the structural parameters, operating parameters and ventilation system parameters and facility costs of the subway tunnel during operation are known, and the importance index of each node is determined. According to the subway tunnel ventilation network diagram and the determined monitoring effectiveness level, a subway tunnel gas monitoring system sensor layout optimization model based on quantitative risk assessment is generated through a maximum coverage model, and the optimal solution of the function is obtained to obtain the relative positions of the gas monitoring system sensors in the subway tunnel during operation, and the layout result is obtained.

[0023] Furthermore, the evaluation formula for determining the importance index of each node in step S2) is:

[0024] R=P×I

[0025] Among them, P is a probability index, specifically referring to the possibility (probability) of a gas leakage incident in a subway tunnel, with a value range of (0, 1). Determining this value is divided into four steps.

[0026] (1) Predict the possibility of occurrence from a subjective and objective perspective, classify and determine the level. A. Historical probability of occurrence: Classify and determine the level based on the frequency of gas leakage accidents in local subway tunnels from the construction period to the operation stage. B. Comprehensive expert evaluation: Under the organization of the Metro Co., Ltd., risk management experts and experienced personnel from different fields participated in the technical analysis and collective discussion to determine the possibility of gas leakage accidents in subway tunnels, classify and determine the results.

[0027] (2) Analyze and determine the probability level. According to the existing standards or the standard analysis generated through discussion, determine the probability level results of A and B. Among them, the scoring in the expert evaluation mainly considers key risk factors such as geological conditions, equipment aging, personnel density, distribution of key equipment, and ventilation conditions.

[0028] (3) Summarize and calculate to determine the probability of occurrence. The overall probability of a gas leak in a subway tunnel is calculated by summing up the scores of each level and dividing the total by the highest possible score. This value reflects the comprehensive judgment based on objective historical data and subjective expert opinions, providing a quantitative probability indicator for risk management.

[0029] (4) According to Table 1 below, determine the probability level of a subway tunnel gas leakage incident.

[0030] Table 1 Probability classification of gas leakage incidents in subway tunnels

[0031]

[0032] I is the consequence impact index, specifically referring to the consequences of a gas outburst in a subway tunnel. The grading range of I is determined by consulting relevant risk accident experts and referring to the description of tunnel gas leakage in the on-site disposal plan for tunnel gas alarm compiled by a subway company. The grading is based on the gas concentration as follows: When the gas concentration is less than 0.3%, it is level 1. At this time, the maintenance personnel strengthen the remote monitoring of the gas concentration in the area. When the gas concentration is between 0.3% and 0.5%, it is level 2. At this time, the dispatching and command center immediately starts the tunnel ventilation equipment according to the gas exceeding standard working condition, and the maintenance unit immediately organizes professionals to conduct on-site inspections (passenger trains can be added or other methods can be used). When the gas concentration is between 0.5% and 1.5%, it is level 3. At this time, the dispatching and command center immediately starts the tunnel ventilation equipment according to the gas exceeding standard working condition, and the dispatching and command center immediately stops the passenger trains heading to the gas leakage area, organizes the passenger trains that have entered the section to return to the departure station, and organizes the passenger trains that cannot be stopped in time to travel to the next station, and the maintenance unit immediately organizes professionals to conduct on-site inspections. When the gas concentration is greater than 1.5%, it is Level 4. The maintenance unit stops the operation of tunnel fans within 100m of the alarm gas sensor and keeps other fans running. The dispatching and command center immediately stops trains heading to the gas leakage area and organizes trains that have entered the section to return to the departure station. If there is no time to stop the trains, they are organized to travel to the next station. The dispatching and command center cuts off the power supply to the contact network within 100m of the alarm gas sensor.

[0033] Finally, by calculating the product of “damage consequences of gas leakage in subway tunnels” and “possibility of gas leakage in subway tunnels” (i.e., R value), the importance values ​​are divided into three levels, as shown in Table 2 below.

[0034] Table 2 Risk matrix of gas leakage in subway tunnels

[0035]

[0036] Furthermore, step S4 specifically includes the following steps:

[0037] SS1, based on the Dantzing-Wolfe decomposition technique (DW decomposition technique), the mathematical model of the original problem is decomposed into the main problem of maximizing the disaster preparedness of the entire area by the sensor layout of the subway tunnel gas monitoring system and the pricing sub-problem;

[0038] SS2. Solve the relaxation of the master problem using the column generation algorithm. The initial solution obtained generates some feasible columns to form a restricted master problem. The restricted master problem is linearly relaxed. The relaxation problem is solved by the linear programming method to obtain the optimal solution of the relaxation problem and the dual variable value.

[0039] SS3, bring the dual variable into the sub-problem and re-price it, add the columns with test numbers less than 0 in the sub-problem to the restricted main problem and iterate again, that is, return to step SS2, until the sub-problem cannot generate a new column with a test number less than 0, which is the final optimal solution;

[0040] SS4. If the linear relaxation is an integer solution, store the solution. If the linear relaxation is not an integer solution, perform a branching judgment. If it does not meet the branching requirements, perform a branching operation and return to step SS2.

[0041] Furthermore, the objective function of the main problem is:

[0042] Indicates that the gas disaster preparedness of the entire region is maximized

[0043] The constraints include:

[0044]

[0045] Where: i is the scheduling strategy of a single gas sensor, describing the feasible sequence of whether a single gas sensor is installed and connected to the gas monitoring system; z i is a binary variable. If a certain point in the subway tunnel is a gas monitoring system sensor deployment point, then z i is 1, otherwise it is 0; i It indicates the ability to accurately prevent gas outburst incidents in the surrounding area when the gas sensor is arranged at point i.

[0046] Furthermore, the objective function of the pricing sub-problem is:

[0047]

[0048] A greedy algorithm is used to solve the maximization pricing subproblem.

[0049] Furthermore, in step SS4, if the linear relaxation is not an integer solution, a branch shearing judgment is performed, and the specific steps are as follows:

[0050] SS5-5-1, if the optimal objective value of the relaxed main problem is not greater than the current optimal integer solution, then prune the branches and end, otherwise execute step S5-5-2;

[0051] S5-5-2, if the optimal solution of the relaxed main problem is an integer solution, replace the current optimal solution with the optimal solution of the relaxed main problem, then prune the branches and end, otherwise execute step S5-5-3;

[0052] S5-5-3. If the optimal objective value of the relaxed master problem is less than the current optimal integer solution, and some variables Z i If the number is a decimal, branch.

[0053] Further, the steps of the branch described in step S5-5-3 are as follows:

[0054] S5-5-3-1. Select branch variables r * For the corresponding installation strategy i * Installation sequence:

[0055] S5-5-3-2, choose any task node i and (j,i)∈r * ;

[0056] S5-5-3-3. On one branch, the arc (j, i) is prohibited from being accessed, and on another branch, all arcs except the arc (j, i) are prohibited from being accessed.

[0057] The present invention has at least the following beneficial effects:

[0058] 1. Accurate monitoring: This method uses advanced sensing technology and data processing algorithms to accurately monitor the gas concentration in different areas of the subway tunnel, significantly improving the accuracy and reliability of the monitoring system in complex environments.

[0059] 2. Cost efficiency: By building a maximum coverage site selection model and introducing a branch pricing optimization algorithm, the number of sensors required is effectively reduced. This optimization not only reduces the initial investment, but also significantly reduces the long-term operating cost of the system by reducing maintenance requirements and energy consumption.

[0060] 3. System flexibility and scalability: The distributed sensor layout strategy enhances the system's coverage of all parts of the subway tunnel and improves the flexibility and scalability of the overall system. This layout facilitates system upgrades and maintenance based on future technological advances or operational needs, ensuring long-term technical adaptability and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only illustrate embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0062] Figure 1 This is a diagram of the layout of traditional fixed-spacing sensors;

[0063] Figure 2 This is a diagram of the first sensor arrangement scheme when the first-level, second-level and third-level risk sources appear simultaneously in the embodiment;

[0064] Figure 3This is a diagram of the sensor arrangement scheme when two primary risk sources appear at the same time in the embodiment;

[0065] Figure 4 1 is a diagram of sensor arrangement when there is one primary risk source and one secondary risk source in the embodiment. DETAILED DESCRIPTION

[0066] The specific implementation modes of the present invention are described below to facilitate those skilled in the art to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0067] The monitoring effectiveness level refers to the shortest response time required for the gas sensor at the downstream node j to detect the gas outburst at the upstream node i in the wind network. According to different gas monitoring levels (effective warning time of the gas monitoring system), the maximum coverage site selection model (MCLP) is used to determine the best sensor layout points. The set of optional gas monitoring points is J, the set of demand points is I, P represents the number of gas sensors, and the set of gas monitoring points that effectively monitors the demand point i is recorded as: N i ={j∈J|c ij =1}, the maximum coverage site selection model is described as follows:

[0068]

[0069] satisfy:

[0070] Where: i is the importance index of each demand node i obtained based on the probability-impact risk assessment model. The objective function is to maximize the disaster preparedness target of the entire region. Formula (2) selects P from the set of candidate monitoring points J. j Gas monitoring points, x j Indicates whether the candidate monitoring point j is selected. If x is selected j =1, otherwise x j =0,y i Indicates whether the gas situation at demand point i can be effectively monitored. If the monitored y i =1, otherwise y i =0; In this model, it is assumed that the cost of installing gas sensors at each alternative layout point is the same. ij is an element in the node effective monitoring matrix or branch effective monitoring matrix at a given monitoring effectiveness level, c ij ∈{0,1}, let cij = 1, it is assumed that the sensor at node i can monitor the gas situation at node j in a timely and effective manner; otherwise, let c ij = 0, it is considered that the sensor at node i cannot monitor the gas situation at node j, and the effective monitoring matrix C = [c ij ] is constructed based on the node monitoring time matrix T.

[0071] The node time monitoring matrix is ​​constructed as follows:

[0072] The node adjacency matrix of the wind network G with respect to the shortest wind flow time between nodes is constructed by multiplying the wind network node set V by itself, which is recorded as: T = [t ij ], the process is as follows:

[0073]

[0074] Where N is the total number of nodes in the wind network; t ij is the wind flow time between node i and node j, in seconds.

[0075] The effective monitoring matrix C = [c ij ]The construction process is as follows:

[0076] Based on the node monitoring time matrix T, an effective monitoring matrix C = [c ij ], where c ij ∈{0,1}, the construction process is as follows: given the monitoring effective level M L , when t ij ≤M L When c ij = 1, it is assumed that the sensor at node i can monitor the gas situation at node j in a timely and effective manner; otherwise, let c uj = 0, it is considered that the sensor at node i cannot monitor the gas situation at node j; therefore, the meaning of the effective monitoring matrix C is: the non-zero elements in the jth column of the matrix C represent the effective level M of the monitoring L Under this condition, node j can effectively monitor node i.

[0077] S4. Further, the steps of the branch pricing algorithm described in S4 are as follows:

[0078] S4-1, based on the Dantzing-Wolfe decomposition technique (DW decomposition technique), the mathematical model of the original problem is decomposed into the main problem of maximizing the disaster preparedness of the entire area by arranging sensors for the subway tunnel gas monitoring system and the pricing sub-problem;

[0079] S4-2, using a column generation algorithm to solve the relaxation of the master problem, the obtained initial solution generates a partially feasible column to form a restricted master problem, the restricted master problem is linearly relaxed, the relaxation problem is solved by a linear programming method, and the optimal solution of the relaxation problem and the dual variable value are obtained;

[0080] S4-3, bring the dual variable into the sub-problem and re-price it, add the columns whose test numbers are less than 0 in the sub-problem to the restricted main problem and iterate again, that is, return to step SS2, until the optimal solution is obtained;

[0081] S4-4. If the linear relaxation is an integer solution, the solution is stored. If the linear relaxation is not an integer solution, a branching judgment is performed. If the branching requirement is not met, a branching operation is performed and the process returns to step SS2.

[0082] The constraints representing the maximum disaster preparedness for the entire region include:

[0083]

[0084] Where: y i is a binary variable. If a certain point in the subway tunnel is a gas monitoring system sensor deployment point, then y i is 1, otherwise it is 0; w i It represents the risk of gas outburst incident when gas sensors are arranged at point i, and represents the importance index of each demand node i. i is the scheduling strategy of a single gas sensor, which describes the feasible installation sequence of a single gas sensor under the constraint of meeting the basic requirements of gas outburst risk monitoring of sensors at each stage of the gas monitoring system;

[0085] In this model, i ∈{0,1}, Change to z w ≥0, The linear relaxed master problem of the master problem is obtained, and the linear relaxed master problem is solved and recorded as the corresponding dual variable η.

[0086] In one embodiment, the objective function of the pricing sub-problem is:

[0087]

[0088] A greedy algorithm is used to solve the maximization pricing subproblem.

[0089] In one embodiment, in step S4-4, if the linear relaxation is not an integer solution, a branch cutting judgment is performed, and the specific steps are as follows:

[0090] S5-5-1, if the optimal objective value of the relaxed master problem is greater than the current optimal integer solution, then prune the branches and end, otherwise execute step S5-5-2;

[0091] S5-5-2, if the optimal solution of the relaxed main problem is an integer solution, replace the current optimal solution with the optimal solution of the relaxed main problem, then prune the branches and end, otherwise execute step S5-5-3;

[0092] S5-5-3. If the optimal objective value of the relaxed main problem is greater than the current optimal integer solution and some variables are decimals, branching is performed.

[0093] In one embodiment, the steps of the branch in step S5-5-3 are as follows:

[0094] S5-5-3-1. Select branch variables r * For the corresponding installation strategy i * Installation sequence:

[0095] S5-5-3-2, choose any task node i and (j,i)∈r * ;

[0096] S5-5-3-3. On one branch, the arc (j, i) is prohibited from being accessed, and on another branch, all arcs except the arc (j, i) are prohibited from being accessed.

[0097] In the above step S4, the determined multiple subway tunnel gas monitoring system sensor layout optimization schemes are simulated by using Vissim software to determine the final best scheme.

[0098] In summary, the sensor layout optimization method of the subway tunnel gas monitoring system based on quantitative risk assessment provided by the present invention collects the structural parameters of the target subway tunnel, the sensor operation and maintenance cost, the facility cost and the assembly cost parameters, and considers the risk indicators such as geological conditions, equipment aging degree, personnel density, key equipment distribution, ventilation conditions, etc. to conduct risk assessment; according to the subway tunnel ventilation network diagram and the determined monitoring effectiveness level, the subway tunnel gas monitoring system sensor layout optimization model based on quantitative risk assessment is generated through the disaster preparedness maximization model, and the model is solved to output the layout plan of the gas sensor facility that takes both economy and safety into consideration. The present invention mainly solves the problem of sensor facility layout optimization of the subway tunnel gas monitoring system, takes into account the safety of subway tunnel construction, and quantitatively couples the risk cost into the total cost for layout optimization research. In addition, the present invention can also reduce the risk of injury to facilities and people when accidents occur while meeting the requirements of the design specifications. Based on the perspective of layout optimization, the present invention proposes a sensor facility layout optimization method of the subway tunnel gas monitoring system based on quantitative risk assessment, which is used to help decision makers create a low-risk layout structure.

[0099] The present invention also provides a method and application for designing and arranging sensors of a distributed gas monitoring system for a subway tunnel with a risk of gas outburst. The selected engineering background is a tunnel through mountains and ridges, which adopts a double-tunnel repair scheme (line spacing of 30m), a mining method construction, a tunnel entrance connected to a "U"-shaped open structure, and an exit connected to a roadbed project. The starting and ending mileages are assumed to be in the range of YK01+285.000 to YK11+000.000, and the total length of the project is 9715m. A single-direction line tunnel of an urban subway with a length of 1000m (the position of the section pile number is assumed to be in the range of YK05+000.000 to YK06+000.000) is selected as the area to be monitored. During the construction of the subway tunnel, the area was a high-gas work area as a whole. According to the analysis and detection of the gas distribution in the tunnel range by gas drilling detection, it can be seen that there are high-gas outburst risk areas and low-gas outburst risk areas within the project range. During the construction period, corresponding safety technical measures were taken to deal with the gas storage area by using methods such as drilling and drainage, grouting and plugging, spray protection and lining. After completion, no gas leakage was found after testing.

[0100] During normal operation of the tunnel, natural ventilation and train piston air exchange are designed. The ventilation direction in the tunnel is consistent with the driving direction, and mechanical ventilation is generally not required. However, when the gas in the tunnel exceeds the standard by 0.3% due to construction quality or defects, the gas monitoring system links the tunnel fan, jet fan and other ventilation equipment to perform mechanical ventilation to dilute the gas.

[0101] In order to ensure the safety of the tunnel during subway operation, a tunnel gas monitoring system is set up in the areas where gas leakage may occur. Gas monitoring sensors are installed on the tunnel walls. The detection range of a single gas sensor is selected to be 100m, so the number of gas sensors installed in the preset area is 10. Therefore, there are optional sensor installation locations such as Figure 1 Shown is the default gas sensor installation.

[0102] According to the sensor layout strategy proposed in the invention content of this patent, a sensor installation plan covering the entire monitoring area can be obtained, wherein the installation plan includes the following steps:

[0103] 1. The evaluation formula for determining the risk index of each node is:

[0104] R=P×I

[0105] Among them, P is the probability index, specifically refers to the probability of a gas outburst incident, and its value range is (0, 1); I is the consequence impact index, specifically refers to the consequence caused by a gas outburst incident, and is determined based on the gas concentration monitored by the node when a leakage occurs in the tunnel.

[0106] 2. Establish the node adjacency matrix T of the subway tunnel wind network

[0107] According to the adjacency matrix construction method, the wind network node set V is multiplied by itself to construct the node adjacency matrix of the wind network G about the shortest wind flow time between nodes, which is recorded as: T = [t ij ], the process is as follows:

[0108]

[0109] Where N is the total number of nodes in the wind network; t ij is the wind flow time between node i and node j, in seconds.

[0110] 3. Establish an effective monitoring matrix C

[0111] Based on the node monitoring time matrix T, an effective monitoring matrix C = [c ij ], the construction process is as follows: according to the risk value obtained in step 1, the risk is classified into levels, where the importance values ​​1, 2, and 3 are classified as first-level risks, 4, 6, and 8 are second-level risks, and 9, 12, and 16 are third-level risks. Different risk levels have different monitoring effective times. Given a monitoring effective level M L , when t ij ≤M L When c ij = 1, it is assumed that the sensor at node i can monitor the gas situation at node j in a timely and effective manner; otherwise, let c ij=0 means that the sensor at node i cannot monitor the gas situation at node j.

[0112] 4. Construct a model to maximize disaster preparedness within the region

[0113] The constraints representing the maximum disaster preparedness for the entire region include:

[0114]

[0115] Where: y i is a binary variable. If a certain point in the subway tunnel is a gas monitoring system sensor deployment point, then y i is 1, otherwise it is 0; w i It represents the risk of a gas outburst event when the gas sensor is arranged at point i, and represents the importance index of each demand node i. i is the scheduling strategy of a single gas sensor, which describes the feasible installation sequence of a single gas sensor under the constraint of meeting the basic requirements of gas outburst risk monitoring of sensors at each stage of the gas monitoring system.

[0116] 5. Branch and price algorithm to solve the disaster preparedness maximization model in the above segment

[0117] (1) Based on the Dantzing-Wolfe decomposition technique (DW decomposition technique), the mathematical model of the original problem is decomposed into the main problem of maximizing the disaster preparedness of the entire area through the sensor layout of the subway tunnel gas monitoring system and the pricing subproblem.

[0118] (2) The column generation algorithm is used to solve the relaxation of the master problem. The initial solution is used to generate a partially feasible column to form a restricted master problem. The restricted master problem is linearly relaxed and solved by a linear programming method to obtain the optimal solution of the relaxation problem and the dual variable value.

[0119] (3) Bring the dual variable into the subproblem and reprice it. Add the columns whose test numbers are less than 0 in the subproblem to the restricted main problem and iterate again, that is, return to step (2) until the optimal solution is obtained.

[0120] (4) If the linear relaxation is an integer solution, the solution is stored. If the linear relaxation is not an integer solution, a branching judgment is performed. If the branching requirement is not met, a branching operation is performed. The specific steps of branching are as follows:

[0121] If the optimal objective value of the relaxed master problem is greater than the current optimal integer solution, then prune the branches and end, otherwise proceed to the next step;

[0122] If the optimal solution of the relaxed main problem is an integer solution, the optimal solution of the relaxed main problem will replace the current optimal solution, and then the branches will be pruned and the process will end. Otherwise, the next step will be executed.

[0123] If the optimal objective value of the relaxed main problem is greater than the current optimal integer solution and some variables are decimals, branching is performed. The branching steps are as follows:

[0124] Select branch variable | Z i* -0.5|=min i |Z i* -0.5|,r * For the corresponding installation strategy i * Installation sequence:

[0125] Choose any task node i and (j,i)∈r * ;

[0126] Let arc (j, i) be prohibited from access on one branch, and all arcs except arc (j, i) be prohibited from access on another branch.

[0127] According to the historical gas leakage accident data of a subway tunnel from the construction period to the operation period, four sections of 1km were randomly selected (all of which had gas leakage of different levels during the tunnel construction period), and four situations of gas leakage hazard source risk level and its distribution were determined based on the actual situation and the expert's score on the gas leakage point. At the same time, from the subway tunnel ventilation data in this embodiment, it can be seen that the tunnel wind speed is determined to be 2m / s, then the first-level risk monitoring effective level is 150s, the second-level risk monitoring effective level is 100s, and the third-level risk monitoring effective level is 50s. Based on the default sensor layout method, the above risk assessment results and basic data such as tunnel wind speed are substituted into the constructed model, and the optimal layout of the gas monitor is obtained through the above-mentioned branch pricing method. The specific optimization layout results of the gas sensor are as follows:

[0128] (a) If Figure 2 As shown in the figure, the tunnel section has level 1, level 2 and level 3 risk sources at the same time. According to the monitoring effectiveness level of each level of risk source, it can be seen that the calculation result meets the monitoring distance requirements for each level of risk source. At the same time, the number of gas monitors arranged in this section is reduced to three.

[0129] (II) There are two primary risk sources in the tunnel section, which are concentrated in the middle of the section. In this case, only two gas monitoring sensors are needed to cover the monitoring needs of the low-risk tunnel section. Figure 3 shown.

[0130] (III) There is one primary risk source and one secondary risk source in each section of the tunnel, which are loosely distributed on both sides of the section. It can be found that only two gas detectors are needed to cover the monitoring needs of the low-risk tunnel section, such as Figure 4 shown.

[0131] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they are aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. It should be pointed out that any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for deploying a gas sensor, characterized in that: The steps include: S1: Collect relevant gas parameters of subway tunnels during operation; The gas-related data include structural parameters, operating parameters and ventilation system parameters of the subway tunnel; S2: Use the probability-impact based risk assessment method to evaluate the risk of each point and determine the importance index of each node; S3: Based on the subway tunnel ventilation network diagram and the determined monitoring effectiveness level, the subway tunnel gas monitoring system sensor layout optimization model based on quantitative risk assessment is generated through the maximum coverage site selection model; In step S3, the sensor layout of the subway tunnel gas monitoring system aims to maximize the disaster preparedness target, and the objective function is: Where: i is the importance index of each demand node i obtained based on the probability-impact risk assessment model. The objective function is to maximize the disaster preparedness target of the entire region. i Indicates whether the gas situation at demand point i can be effectively monitored. If the monitored y i =1, otherwise y i =0; The constraints of step S3 are as follows: satisfy: Where: Formula (2) represents the selection of P from the candidate monitoring point set J j Gas monitoring points, x j Indicates whether the candidate monitoring point j is selected. If x is selected j =1, otherwise x j =0,y i Indicates whether the gas situation at demand point i can be effectively monitored. If the monitored y i =1, otherwise y i =0; In this model, it is assumed that the cost of installing gas sensors at each alternative layout point is the same; c ij is an element in the node effective monitoring matrix or branch effective monitoring matrix at a given monitoring effectiveness level, c ij ∈{0,1}, let c ij = 1, it is assumed that the sensor at node i can monitor the gas situation at node j in a timely and effective manner; otherwise, let c ij = 0, it is considered that the sensor at node i cannot monitor the gas situation at node j, and the effective monitoring matrix C = [c ij ] is constructed based on the node monitoring time matrix T; S4: Solving the optimization model described in step S3 based on the branch pricing algorithm to obtain the optimal gas sensor layout strategy, wherein the gas sensor layout strategy in the subway tunnel includes determining the gas sensor deployment points.

2. A method for deploying a gas sensor according to claim 1, characterized in that: In step S1, the optimization of the sensor layout of the gas monitoring system in the subway tunnel based on quantitative risk assessment is to conduct risk assessment on each point using a probability-impact based risk assessment method under the condition that the structural parameters, operating parameters, ventilation system parameters and facility costs of the subway tunnel during operation are known, determine the importance index of each node, and generate a subway tunnel gas monitoring system sensor layout optimization model based on quantitative risk assessment through a maximum coverage model according to the subway tunnel ventilation network diagram and the determined monitoring effectiveness level, and find the optimal solution to the function, obtain the relative positions of the gas monitoring system sensors in the subway tunnel during operation, and find the layout result.

3. The method for deploying a gas sensor according to claim 1, characterized in that: The evaluation formula for determining the importance index of each node in step S2 is: R=P×I Among them, P is a probability index, specifically referring to the possibility of a gas leakage event in a subway tunnel, with a value range of (0, 1). Determining this value is divided into four steps; (1) Predict the possibility of occurrence from a subjective and objective perspective, classify and determine the level; A. Historical probability of occurrence: classify and determine the level based on the frequency of gas leakage accidents in local subway tunnels from the construction period to the operation stage. B. Comprehensive expert evaluation: Under the organization of the Metro Co., Ltd., risk management experts and experienced personnel from different fields participated in the technical analysis and collective discussion to determine the possibility of gas leakage accidents in subway tunnels, classify and determine the results; (2) Analyze and determine the probability level; determine the probability level A and B results based on existing standards or standards generated through discussion; the expert evaluation mainly considers geological conditions, equipment aging, personnel density, key equipment distribution, and ventilation conditions as key risk factors; (3) Summarize and calculate to determine the probability of occurrence. The overall probability of a gas leak in a subway tunnel is calculated by summing up the scores of each level and dividing the total by the highest possible score. This value reflects the comprehensive judgment based on objective historical data and subjective expert opinions, and provides a quantitative probability indicator for risk management. (4) The likelihood of a gas leakage incident in a subway tunnel is graded. The importance value is divided into three levels by calculating the product of "damage consequences of a gas leakage in a subway tunnel" and "possibility of a gas leakage in a subway tunnel".

4. The method for deploying a gas sensor according to claim 1, characterized in that: Step S4 specifically includes the following steps: SS1, based on the Dantzing-Wolfe decomposition technique (DW decomposition technique), the mathematical model of the original problem is decomposed into the main problem of maximizing the disaster preparedness of the entire area by the sensor layout of the subway tunnel gas monitoring system and the pricing sub-problem; SS2. Solve the relaxation of the master problem using the column generation algorithm. The initial solution obtained generates some feasible columns to form a restricted master problem. The restricted master problem is linearly relaxed. The relaxation problem is solved by the linear programming method to obtain the optimal solution of the relaxation problem and the dual variable value. SS3, bring the dual variable into the sub-problem and re-price it, add the columns with test numbers less than 0 in the sub-problem to the restricted main problem and iterate again, that is, return to step SS2, until the sub-problem cannot generate a new column with a test number less than 0, which is the final optimal solution; SS4. If the linear relaxation is an integer solution, store the solution. If the linear relaxation is not an integer solution, perform a branching judgment. If it does not meet the branching requirements, perform a branching operation and return to step SS2.

5. A method for deploying a gas sensor according to claim 4, characterized in that: The objective function of the main problem is: The constraints representing the maximum gas disaster preparedness level of the entire region include: Where: i is the scheduling strategy of a single gas sensor, describing the feasible sequence of whether a single gas sensor is installed and connected to the gas monitoring system; z i is a binary variable. If a certain point in the subway tunnel is a gas monitoring system sensor deployment point, then z i is 1, otherwise it is 0; i It indicates the ability to accurately prevent gas outburst incidents in the surrounding area when the gas sensor is arranged at point i.

6. A method for deploying a gas sensor according to claim 4, characterized in that: The objective function of the pricing subproblem is: A greedy algorithm is used to solve the maximization pricing subproblem.

7. The method for deploying a gas sensor according to claim 4, characterized in that: In step SS4, if the linear relaxation is not an integer solution, a branch shearing judgment is performed, and the specific steps are as follows: SS5-5-1, if the optimal objective value of the relaxed main problem is not greater than the current optimal integer solution, then prune the branches and end, otherwise execute step S5-5-2; S5-5-2, if the optimal solution of the relaxed main problem is an integer solution, replace the current optimal solution with the optimal solution of the relaxed main problem, then prune the branches and end, otherwise execute step S5-5-3; S5-5-3. If the optimal objective value of the relaxed master problem is less than the current optimal integer solution, and some variables Z i If the number is a decimal, branch.

8. A method for deploying a gas sensor according to claim 7, characterized in that: The steps of the branch described in step S5-5-3 are as follows: S5-5-3-1. Select branch variable | Z i* -0.5|=min i |Z i* -0.5|,r * For the corresponding installation strategy i * Installation sequence: S5-5-3-2, choose any task node i and (j,i)∈r * ; S5-5-3-3. On one branch, the arc (j, i) is prohibited from being accessed, and on another branch, all arcs except the arc (j, i) are prohibited from being accessed.

Images