Subway entrance and exit flood control equipment quick response method based on water level situation analysis

Abstract

The invention discloses a metro entrance and exit flood control equipment quick response method based on water level situation analysis, which comprises the following steps: acquiring water level data of different metro entrances and exits through monitoring video images, correcting image distortion by adopting inverse transparent transformation, extracting a water level line position by applying a threshold segmentation technology, and ensuring the monitoring precision; the water level change trend is predicted by combining a seasonal gating circulation unit neural network, a flood control measure model based on a decision-making tree is generated, rapid response of flood control equipment is achieved, and the bidirectional Bayesian decision-making model is used for evaluating the risk of subway entrance and exit flood disasters and making corresponding prevention and control measures. The real-time performance and accuracy of water level situation analysis and the response efficiency of flood control measures are improved, and the method has the advantages that the water level situation at the entrance and the exit of the subway can be accurately monitored in real time, different water level conditions are quickly responded through the decision-making tree model, and therefore the influence of flood on a subway system is effectively reduced, and the safety of passengers and facilities is guaranteed.

Description

Technical Field

[0001] The present invention relates to the fields of deep learning and artificial intelligence, and in particular to a rapid response method for flood control equipment at subway entrances and exits based on water level situation analysis. Background Art

[0002] It is particularly important to design and implement a rapid response method for subway entrance and exit flood control equipment based on water level situation analysis. In recent years, my country's urbanization process has accelerated. As an important part of urban transportation, the safe operation of subway is of vital importance. However, the problem of subway entrance and exit flood control has always been a shortcoming of subway safety operation. In order to respond to subway entrance and exit flood disasters more quickly, a more timely and effective solution is needed.

[0003] In recent years, with the rapid development of machine learning and artificial intelligence technology, especially the application of deep learning in images, new ideas and methods have been provided for water level situation analysis. As an emerging deep learning model, the gated recurrent unit neural network (GRU) has demonstrated its powerful time series processing capabilities in traffic flow forecasting, weather forecasting and other fields. The Bow-tie model is a risk management tool that analyzes and displays the causes and consequences of accidents in a graphical way, as well as how to control risks through preventive measures and emergency responses. By capturing the temporal correlation of data, GRU can effectively handle the temporal dependency in sequence data, thus providing the possibility for dynamic and complex data analysis. GRU can update the model state in real time, respond quickly to the latest data changes, and provide timely data forecasts. Applying GRU to the prediction of subway entrance and exit water levels in combination with seasonal attributes can take advantage of its advantages and real-time performance in processing time series data to capture risk signals from multiple dimensions such as historical water level data and seasonal characteristics. Specifically, the seasonal cycle prediction model can conduct in-depth analysis of the water level data at subway entrances and exits, and identify potential flood risk factors and risk propagation paths. By learning the complex interaction of water level data in time and space, the seasonal cycle prediction model can predict the water level change trend of subway entrances and exits in the future. Combined with the Bow-tie model, it provides decision support for the subway operation management and helps the subway system to formulate more effective flood control management strategies and countermeasures. In addition, the subway entrance and exit water level prediction method based on the seasonal cycle prediction model also has strong generalization ability and adaptability. It can process subway entrance and exit water level data from different regions and different climatic conditions, and even adapt to the needs of subway systems of different sizes. By continuously learning meteorological changes and water level dynamics at subway entrances and exits, the seasonal cycle prediction model can continuously optimize the prediction model and improve the accuracy and reliability of water level prediction. The subway entrance and exit water level prediction method based on the seasonal cycle prediction model is not only a technical innovation, but also an inevitable product to adapt to the trend of urban waterlogging control and smart city construction. As more and more cities begin to pay attention to and apply this method, the rapid response of subway entrance and exit flood control equipment based on water level situation analysis is expected to become an important tool for flood control management of subway systems, providing strong support for the safe operation of subways. Through the advanced prediction capabilities of the seasonal cycle prediction model, water level risks at subway entrances and exits are effectively monitored, thereby ensuring the travel safety of citizens and improving the overall disaster resistance of the subway system. Summary of the invention

[0004] Traditional methods often rely on manual judgment and operation, resulting in slow response when floods come, and unable to effectively prevent flood backflow in a timely manner. Moreover, traditional methods only take action when a flood has occurred or is about to occur, lacking advance prediction and early warning of water level conditions, and unable to prevent disasters before they occur. In view of this, the purpose of the present invention is to provide a rapid response method for subway entrance and exit flood control equipment based on water level situation analysis to solve the problem that traditional subway entrances and exits have slow response speed to water level conditions and lack of predictability.

[0005] The objective of the present invention is achieved through the following technical solutions:

[0006] The method for rapid response of subway entrance and exit flood control equipment based on water level situation analysis includes the following steps:

[0007] Step S1: Collect and obtain surveillance video images of different subway entrances and exits, read the video images frame by frame, and obtain a water level image sequence at different times;

[0008] Step S2: applying the inverse perspective transformation method of point-to-point control to correct the distorted image;

[0009] Step S3: performing target area extraction and threshold segmentation operations on the corrected water level image;

[0010] Step S4: extract the water level position of each frame from the processed image sequence, and analyze the water level situation according to the position change of the water level;

[0011] Step S5: Generate a flood control decision tree model based on the water level situation, which is divided into different risk levels, and the flood control equipment takes corresponding flood control measures according to the level.

[0012] Further, the step S1 specifically includes:

[0013] Step S101: Acquire surveillance video images of multiple subway entrances and exits;

[0014] Step S102: reading the collected surveillance videos of subway entrances and exits frame by frame using a dynamic synchronous frame extraction method;

[0015] Step S103: dividing the read video image into water level image sequences at different moments according to time;

[0016] Further, the step S2 specifically includes:

[0017] Step S201: randomly determine four points in the three-dimensional space to record the position coordinates of four feature points in the water level image;

[0018] Step S202: Based on the position coordinates determined in S201, specify the real position coordinates under the normal viewing angle, and calculate the corresponding relationship between the two sets of position coordinates;

[0019] Step S203: converting the monitoring imaging coordinate system into an abstract pixel coordinate system;

[0020] The specific method of converting the pixel coordinate system in step S203 is as follows:

[0021] First, select the four pixels at the outermost edge of the image as feature points, and record the coordinates of these four control points, denoted as (x1, y1), (x2, y2), (x3, y3), (x4, y4), and use their corresponding actual coordinates (u1, v1), (u2, v2), (u3, v3), (u4, v4) as known quantities; use perspective transformation to transform the target image into another plane image, and the transformation process is calculated using the following formula:

[0022]

[0023] Among them, (x, y) is the pixel coordinates of the perspective distorted image, (u, v) is the pixel coordinates of the corrected image, a, b, c, d, e, f, p, q are perspective transformation parameters. For the convenience of calculation, the mathematical formula needs to be converted into a matrix form to obtain the corresponding perspective transformation matrix. The rewritten matrix form is as follows:

[0024]

[0025] Then, the coordinates of the four sets of feature points are substituted into the matrix formula to obtain:

[0026]

[0027] Let this matrix be:

[0028] UV=A×M

[0029] According to the above formula, the perspective transformation matrix can be obtained as:

[0030] M=A -1 UV

[0031] Finally, the obtained perspective transformation matrix M is used to process each pixel in the target image to obtain the corrected water level image.

[0032] Further, the step S3 comprises:

[0033] Step S301: manually calibrate the coordinates of the four corner points of the backboard above the water surface based on the corrected water level image. The trapezoidal area enclosed by them is the backboard part of the water level gauge above the water surface. The two waistlines of the trapezoidal area are respectively extended longitudinally to the bottom of the image to obtain the complete backboard area;

[0034] Step S302: Based on the complete backplane area obtained in step S301, the image in RGB format is converted into HSV color space to obtain an image in HSV format;

[0035] The specific method of converting the RGB format image to the HSV format in step S302 is as follows:

[0036] First, normalize the RGB values ​​to the range of 0 to 1 by dividing the value of each color by 255. The specific formula is as follows:

[0037] R n =R / 255,G n =G / 255,B n =B / 255

[0038] Find the maximum value C among the three normalized values max =max(R n ,G n ,B n ) and minimum value C min =min(R n ,G n ,B n ), according to C max and C min Calculate the hue H, saturation S, and brightness V. The calculation method of H, S, and V depends on which color channel contains the maximum value. The specific calculation method is as follows:

[0039] V=C max

[0040]

[0041] Step S303: performing color threshold segmentation and positioning on the water level image in HSV format;

[0042] The method of performing color threshold segmentation and positioning on the water level image in step S303 is as follows:

[0043] First, based on the characteristics of the water level image that are different from other background colors, the image after the target area is extracted is threshold segmented based on red to obtain a binary image M(x,y); the red range is fuzzy processed as: H[0-10&156-180], S[43-255], V[46-255], and the image is segmented according to the threshold of the H channel by setting the HSV three channels to obtain a binary image containing all the red spot areas. The threshold segmentation method for the red area is as follows:

[0044]

[0045] Since the red light spots reflected by the subway entrances and exits are relatively shallow, in order to accurately identify the light spots, the range of the S channel is appropriately widened;

[0046] Then, the m×n neighborhood average calculation method is used to calculate the mean of the binary image M(x,y) to reduce the small red part in the background and obtain the final binary image.

[0047]

[0048] Finally, for the image Perform horizontal projection and vertical projection to determine the position of the water level line in the image;

[0049] Further, the step S4 comprises:

[0050] Step S401: extracting target light spots with changing image background from the water level image processed in S3, using the target tracking algorithm based on adaptive mean shift enhancement to calculate and identify, extracting the number, size, shape, and movement information of the light spots in the image sequence, and obtaining the time series of water levels at different subway entrances and exits;

[0051] The specific steps of the adaptive mean shift enhanced target tracking algorithm in step S401 are as follows:

[0052] First, select the target area to be tracked, and use the target area detected by the last target area matching. The target area should meet the requirement of completely enclosing the target object. Use m grayscale histograms to describe the target model. The target template q μ The probability density of can be described as:

[0053]

[0054] Where μ is the color index of the histogram; m is the number of pixels in the target model; η is the normalization coefficient, so that k(·) is the kernel function. The smaller the distance to the target area, the larger the corresponding pixel weight. h represents the kernel function bandwidth. The larger its value, the smoother the function and the smaller the gradient. x0 is the center of the rectangular box area. i represents the i-th pixel position; f(x i ) represents x i The pixel value corresponding to the pixel position, that is, the quantization sequence number of the pixel point in the quantization feature space; is the matching weight function with the target model;

[0055] Assuming that the size and shape of the target object in each frame of the water level image sequence follow the same rules of the same direction characteristics, let y = (y 1 ,y 2 ) T , x i =(x i 1 ,x i 2 ) T represents the coordinates of the pixel points, N represents the number of pixels, and the pixel points should satisfy:

[0056]

[0057] Then the probability density function q μ It can be expressed as:

[0058]

[0059] Wherein, μ=1,2,3,…,m,η is the normalized coefficient, k(·) is the kernel function; a is the semi-major axis length of the elliptical region of the detected image, and b is the semi-minor axis length of the elliptical region of the detected image;

[0060] Then, in the frame image to be detected, the coordinates of the pixels near the matching position y are expressed as x i Indicates that, where i = 1, 2, 3, ..., N, the kernel function bandwidth is h, μ = 1, 2, 3, ..., m, then the probability density function of the image in the elliptical area is:

[0061]

[0062] Among them, h is the kernel function bandwidth, m represents the number of pixels of the target model in the elliptical area, and η is calculated using Riemann integral h The approximate value of

[0063] Then, the Hellinger distance is used to calculate the model similarity in the above two steps. After adding the obtained coefficient ρ(y, h) to the scale estimation, the similarity between the image to be detected and the target image is expressed as:

[0064]

[0065] For any kernel function bandwidth h0 and h1, they should satisfy:

[0066]

[0067] Finally, in order to obtain the maximum value of ρ(y,h), the partial derivatives of y and h are calculated for ρ(y,h). Thus, the target position in the current frame is and scale factor h1 are:

[0068]

[0069] Among them, the weight w of the pixel value in the candidate model in the current frame i , the weight G of the pixel value of the objective function and the position m of the target image at the kernel function bandwidth h0 k The expression of (y0,h0) is as follows:

[0070]

[0071] w i Use w i bg Replace, it can be expressed as:

[0072]

[0073] Among them, w i bg Represents the weight of the background histogram of pixels near the target area;

[0074] In order to make the described model to be detected better describe the changed part of the shape of the target object, the algorithm adds two regular terms re and rf to the description of the model to be detected. Re affects the prior assumptions, and the formula is expressed as:

[0075]

[0076] Where ζ is the scaling factor;

[0077] rf means that the search window contains background pixels, and the formula is:

[0078]

[0079] Where y and ζ represent the target position and scaling factor respectively, σ is the percentage of weighted background pixels contained in the search window, and the definition formula of B(y,h) is:

[0080]

[0081] In the image to be detected, position yi The calculation scale is g, if |log(h)|>θ s , then calculate the position y in the previous frame image i The target scale is denoted as g b ; if |log(g*g b )|>θ s , then the scale s t for:

[0082] s t =(1-α-β)s t-1 +α*s0+β*h*s t-1

[0083]

[0084] If |log(g)|<θ s ,but:

[0085] s t =(1-γ)*s t-1 +γ*h*s t-1

[0086] Among them, s t-1 Represents y t-1 The scale of the position, s0 represents the default scale, s t Represents y t The scale of the location;

[0087] Step S402: Check the stability of the water level data at the subway inlet and outlet, perform seasonal difference processing on the unstable monthly water level data, and design a seasonal change periodicity prediction model. The model consists of seasonal autoregressive values, seasonal difference values, and error terms. The autoregressive part introduces seasonal parameters to obtain seasonal autoregressive values:

[0088]

[0089] Among them, σ(A) describes the relationship between the water level data and its own hysteresis value; σ i is the coefficient corresponding to the autoregressive order n, A i is an operator that lags by i time units and is used to represent time lag; σ i *A i It represents the relationship between water level data and its data lagged by i time units;

[0090] Introducing seasonal parameters into the difference part can obtain seasonal difference values ​​(1-A s ) d :

[0091]

[0092] Among them, (1-A) d It is used to convert non-stationary time series into stationary time series. d represents the difference order, that is, the number of times the water level data is differentiated; (1-A s ) d It is used to convert non-stationary time series with seasonal characteristics into stationary time series. d is the seasonal difference order, that is, the number of times the water level data is seasonally differenciated; s is the seasonal parameter;

[0093] The established seasonal change periodicity prediction model Ω is:

[0094] Ω=σ(A)φ(A s )(1-A) d (1-A s ) d y i =c+σ(A)*α i

[0095] Among them, c is a constant term, which is used to represent the average trend of water level; α i is the error term, which is used to represent the random fluctuation of water level;

[0096] Step S403: Establish a seasonal cycle prediction model Ψ, and calculate the error sum of squares e of the seasonal change periodicity prediction model Ω and the seasonal cycle prediction model Ψ respectively. j , based on the principle of minimizing the overall error sum of squares, the weights of each prediction model are assigned, and the combined model for subway entrance and exit water level prediction is obtained; the combined weight of the jth model is:

[0097]

[0098] The Ψ model established in step S403 consists of three parts: a season reset gate, a season update gate and a season hidden state. The specific method is as follows:

[0099] First, the Season Update Gate Z t It is determined by the state value of the previous moment and the input value of the current moment. When the value of the seasonal update gate is larger, it means that the current input neuron needs to retain more information, and the information of the previous moment is retained less. The specific expression is as follows:

[0100] Z t =σ(w xz *x t +w hz *h t-1 +b z )

[0101] Among them, x t is the input sequence, h tis the state of the hidden layer, h t-1 is the state of the hidden layer in the last second, w xz is the weight matrix from the input layer to the seasonal update gate, w hz represents the weight matrix of the recurrent connection, σ(·) represents the Sigmoid function, b z Update the gate bias for the seasons, which is used to adapt to different input data by adjusting the threshold of the activation function. It also improves the gradient flow in back propagation, helps avoid network symmetry, and allows each neuron to learn unique features, thereby improving the network's representation ability and training effect;

[0102] Then, the season resets the gate t It is determined by the state value at the previous moment and the input gate at the current moment. When the value of the reset gate is larger, it means that more information is retained at the previous moment. The specific calculation formula is as follows:

[0103] r t =σ(w xr *x t +w hr *h t-1 +b r )

[0104] Among them, w xr is the weight matrix of the input layer to the seasonal reset gate, w hr represents the weight matrix of the recurrent connection, σ(·) represents the Sigmoid function;

[0105] Finally, the output o of the memory unit can be obtained t for:

[0106]

[0107] Among them, c t is the gating signal, h t is the hidden layer state, w xc is the weight matrix from the input layer to the candidate state, w hc represents the weight matrix of the recurrent connection, represents element-wise multiplication, represents the Tanh function;

[0108] Step S404: inputting historical subway entrance and exit water level data into the established combined model, and obtaining future water level data according to seasonal attributes for subsequent flood control equipment analysis;

[0109] Further, the step S5 comprises:

[0110] Step S501: define a risk quantification formula based on the future water level data predicted in S4 to quantitatively assess the risk of flooding. The formula is as follows:

[0111] R(E i )=B*P(E i ) α *(I(E i )*V(E i )) β

[0112] Among them, E i is the flood event to be calculated; P(E i ) is event E i Probability of occurrence; I(E i ) is event E i The degree of impact when it occurs; V(E i ) is event E i The vulnerability of the model to the event; B is the risk adjustment factor, which can be specifically expressed as: B = 1 + ε, where ε is the risk preference coefficient; α and β are weight factors for the probability and degree of impact, which are obtained by studying the probability and degree of impact of flood events at subway entrances and exits;

[0113] Step S502: define four risk levels of low risk, medium risk, high risk and extremely high risk, and set a two-way Bayesian decision model for each entrance and exit to analyze the cause of the accident and the possible harm, so that the model can generate preventive measures before the accident and control measures after the accident.

[0114] The bidirectional Bayesian risk model established in step S502 is a risk assessment model based on the Bow-Tie-Bayesian network and is specifically as follows:

[0115] First, construct an accident tree. The specific method is as follows: determine the upper-level events, intermediate events, and lower-level events, and use logic gates to connect the upper and lower-level events. If all lower-level events must occur at the same time to cause the upper-level event to occur, use an AND gate to connect; conversely, if only one or some of the lower-level events can cause the upper-level event to occur, use an OR gate to connect;

[0116] Then, the event tree is constructed, specifically determining the initial event, determining the safety function and drawing the event tree;

[0117] Furthermore, the structural importance in the bidirectional Bayesian decision model is calculated. By calculating the number of times each lower-level event appears in the minimum cut set, the structural importance of each lower-level event is determined, and the minimum cut set of the fault tree is obtained, which is the minimum combination of events that lead to the occurrence of the upper-level event. Structural importance analysis can evaluate the reliability and safety of the model and help determine which safety measures to prioritize to improve the safety of the model. The calculation formula of structural importance is as follows:

[0118] Among them, t i ∈r i

[0119] Among them, q represents the total number of minimum cut sets, r i represents the i-th minimum cut set, t i represents the number of lower-level events of the i-th minimum cut set, and n represents the number of minimum cut sets;

[0120] Finally, calculate the probability of the upper-level event of the model for the flood control equipment to respond quickly. The specific method is as follows: Replace all events in the original event tree with the events in the minimum cut set to obtain a new equivalent tree. In the equivalent tree, the upper-level event is equal to the union of the minimum cut sets, that is, all events in all minimum cut sets must occur simultaneously to cause the upper-level event to occur. Given that there are r minimum cut sets in the fault tree: E1, E2,..., E k 、E r , then the upper-level event is expressed as:

[0121]

[0122] The probability of the upper-level event occurring is expressed as:

[0123]

[0124] Among them, p i represents the probability of the lower-level event i occurring, k, l, r represent the serial numbers of the minimum cut sets, and the size relationship of the serial numbers is k < l < r; i ≤ k ≤ l ≤ r represents the combination order of the k-th and l-th cut sets in the minimum cut set, x i ∈E k represents the i-th lower-level event belonging to the k-th minimum cut set.

[0125] The beneficial effects of the present invention include:

[0126] (1) By collecting the monitoring video images of the subway entrances and exits, reading and extracting the water level image sequence frame by frame, and combining the inverse perspective transformation technology based on point-to-point control, the present invention accurately corrects the perspective distortion of the images. This technology ensures the accurate identification of the water level line in the images, thereby providing reliable data support for the rapid response of flood control equipment. Compared with the traditional method relying on manual judgment, the present invention significantly improves the automation degree and recognition accuracy of water level monitoring;

[0127] (2) In the preprocessing stage, through the image threshold segmentation technology, the present invention can accurately separate the water level line and the background and other interfering elements, effectively eliminate the interference of noise information, and improve the accuracy of water level line position recognition. Combined with the dynamic monitoring of water level changes, the present invention can effectively avoid the errors that occur when traditional water level monitoring equipment identifies water levels in complex environments. In addition, the use of advanced machine learning algorithms (i.e., adaptive mean shift enhanced target tracking algorithm) can further optimize the image processing process, so that the system can improve the automated response speed and efficiency of subway flood control equipment while continuously learning and adapting to environmental changes;

[0128] (3) The present invention introduces the GRU (Gated Recurrent Unit Neural Network) model, which has powerful time series prediction capabilities through deep learning analysis of historical water level data and real-time meteorological information. The GRU model can effectively capture the time correlation of water level data, update the water level situation in real time, and accurately predict future water level trends. In addition, GRU has a self-learning function, can dynamically optimize the prediction model according to meteorological changes and historical data, has adaptability in different regions and climate conditions, and improves the system's ability to respond to a variety of extreme weather conditions;

[0129] (4) The present invention uses a bidirectional Bayesian network risk assessment model combined with water level situation data to achieve quantitative analysis and graded warning of flood control risks at subway entrances and exits. The model can accurately analyze and estimate the potential risk levels of subway entrances and exits (such as low risk, medium risk, high risk and extremely high risk), and formulate corresponding flood prevention measures by accurately identifying the causes of accidents. This data-driven risk management method can not only improve the safety of the subway system under extreme weather conditions, but also provide a reference for subsequent urban flood control systems, further strengthening the overall disaster resistance of urban infrastructure;

[0130] (5) The present invention realizes a highly integrated intelligent flood control equipment system through the combination of automated monitoring, intelligent analysis and rapid response, which can operate independently without human supervision. When a flood comes, the system can quickly analyze the water level changes and automatically start the corresponding flood control equipment according to the risk level, significantly reducing the lag and error risks caused by human intervention, and providing a strong guarantee for the safe operation of the subway system.

[0131] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description and the preceding claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0132] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0133] Figure 1 This is a flow chart of a method for rapid response of subway exit flood control equipment based on water level situation analysis of the present invention;

[0134] Figure 2 The subway water level map after image processing and threshold segmentation operations in an embodiment of the present invention;

[0135] Figure 3 A bidirectional Bayesian network risk assessment model diagram established for an embodiment of the present invention;

[0136] Figure 4 An event tree diagram of the water level situation at a subway entrance and exit constructed for an embodiment of the present invention;

[0137] Figure 5 The subway entrance and exit water level situation accident tree diagram constructed for the embodiment of the present invention;

[0138] Figure 6 This is a subway water level identification table according to an embodiment of the present invention;

[0139] Figure 7 This is a table of the meanings of the various events in the event tree of an embodiment of the present invention. DETAILED DESCRIPTION

[0140] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the preferred embodiments are only for illustrating the present invention, rather than for limiting the protection scope of the present invention.

[0141] The invention provides a rapid response method for subway entrance and exit flood control equipment based on subway water level situation analysis, such as Figure 1 As shown, the specific steps of the above method will be further described below through a specific embodiment.

[0142] Step S1: Collect and obtain surveillance video images of different subway entrances and exits, read the video images frame by frame, and obtain a water level image sequence at different times;

[0143] Step S1 specifically includes the following sub-steps:

[0144] Step S101: Acquire surveillance video images of multiple subway entrances and exits;

[0145] Step S102: Read the video images frame by frame using a dynamic synchronization frame extraction method for the collected surveillance videos of subway entrances and exits;

[0146] Step S103: dividing the read video image into water level image sequences at different moments according to time;

[0147] Step S2: applying an inverse perspective transformation method based on point-to-point control to correct the distorted image;

[0148] Step S2 specifically includes the following sub-steps:

[0149] Step S201: Randomly determine four points in three-dimensional space to record the position coordinates of the four feature points in the water level image (x1, y1), (x2, y2), (x3, y3), (x4, y4), and the specific corresponding coordinate values ​​are (100, 100), (200, 100), (100, 200), (200, 200); the corresponding actual coordinates (u1, v1), (u2, v2), (u3, v3), (u4, v4) have specific coordinate values ​​of (0, 0), (1, 0), (0, 1), (1, 1), and are brought into the perspective transformation formula to transform the target image into another plane image. The conversion process is calculated using a mathematical formula as shown below:

[0150]

[0151]

[0152] Where (x, y) is the pixel coordinate of the perspective distorted image, (u, v) is the pixel coordinate of the corrected image, a, b, c, d, e, f, p, q are perspective transformation parameters, and the calculated values ​​are a=100, b=0, c=0, d=100, e=0, f=0, p=0.5, q=0.5;

[0153] To facilitate calculation, the mathematical formula needs to be converted into a matrix form to obtain the corresponding perspective transformation matrix. The rewritten matrix form is as follows:

[0154]

[0155] Substitute the coordinates of the four sets of feature points into the matrix formula to obtain:

[0156]

[0157] Let this matrix be:

[0158] UV=A×M

[0159] That is, the perspective transformation matrix is:

[0160] M=A -1 UV

[0161] The inverse transformation matrix is ​​calculated as follows:

[0162]

[0163] Finally, the obtained perspective transformation matrix M is used to process each pixel in the target image to obtain the corrected water level image.

[0164] Step S3: performing target area extraction and threshold segmentation operations on the corrected water level image;

[0165] The image obtained after target extraction and threshold segmentation in this embodiment is as follows Figure 2 As shown, the following operations are included: converting the original image into an HSV format spot image, field average filtering, morphological processing, and threshold segmentation.

[0166] Step S3 specifically includes the following sub-steps:

[0167] Step S301: manually calibrate the coordinates of the four corner points of the backboard above the water surface based on the corrected water level image. The trapezoidal area enclosed by them is the backboard part above the water surface. The two waistlines of the trapezoidal area are respectively extended longitudinally to the bottom of the image to obtain the complete backboard area.

[0168] Step S302: Based on the complete backplane area obtained in step S301, the image in RGB format is converted into HSV color space to obtain an image in HSV format;

[0169] Step S303: performing color threshold segmentation and positioning on the water level image in HSV format;

[0170] Step S4: extract the water level position of each frame from the processed image sequence. Analyze the water level situation according to the position change of the water level;

[0171] The water level prediction in this embodiment is as follows Figure 6 As shown,

[0172] Step S4 specifically includes the following sub-steps:

[0173] Step S401: extracting target light spots with changing image background from the water level image processed in S3, calculating and identifying them using a target tracking algorithm based on adaptive mean shift enhancement, extracting the number, size, shape, and movement information of the light spots in the image sequence, and obtaining a time series of water levels at different subway entrances and exits;

[0174] First, calculate the target model probability density q μ and the candidate model probability density p μ ; In this embodiment, the number of pixels of the target model is N = 1, the kernel function bandwidth is h = 5, the normalization coefficient is C = 1, and the regularization term parameters (re, rf, ζ) = (0.1, 0.2, 0.05); the center pixel coordinates of the target model are (x0, y0) = (100, 100); the pixel coordinates and eigenvalues ​​(x i,b i )=(101,0.8),(102,0.7),(103,0.9),…,(110,0.6); target model size (a,b)=(10,10); candidate model center pixel coordinates (x1,y1)=(110,110); pixel point coordinates and eigenvalues ​​(x i ,b i )=(111,0.7),(112,0.8),(113,0.9),…,(120,0.5); the target model density value and candidate model density value calculated by the candidate model are shown in the following table:

[0175]

[0176] Then, the similarity between the models is calculated according to the formula:

[0177]

[0178] The similarity between the target model and the candidate model is calculated as ρ[ρ(y,h),q]=0.4113;

[0179] Furthermore, the target position and scale are adjusted according to the Hellinger distance ρ. The target position of the current frame is y0 = (101, 100), and the target position is adjusted to y1 = (101.4113, 100.4113) according to m0(y0, h0). According to the initial scale h = 0.5, the scale of the target position is h1 = 0.9113. Finally, the above steps are repeated until y1 converges.

[0180] Finally, the processed water level image sequence is input into the combined water level prediction model, and the seasonal weight matrix is ​​assigned according to the time in the water level image sequence. Specifically, the size of the seasonal weight matrix is ​​set to 0.5 during the flood season (such as July, August, and September). <w x <1, the size of the seasonal weight matrix is ​​set to 0 during the dry season (such as November-February) <w x <0.5, through the combined model prediction, the future water level data is obtained for the rapid response of flood control equipment.

[0181] Step S5: Generate a flood control decision tree model based on the water level situation, which is divided into different risk levels, and the flood control equipment takes corresponding flood control measures according to the level;

[0182] The decision tree model of the associated water level situation and the corresponding flood control measures generated in this embodiment is as follows: Figure 3 Shown

[0183] Step S5 specifically includes the following steps:

[0184] Step S501: define a risk quantification formula based on the future water level data predicted in S4 to quantitatively assess the risk;

[0185] Step S501 shows that the risk quantification formula is defined according to the future water level data predicted by S4 to quantitatively assess the risk. The specific method is as follows:

[0186] First, construct the event tree of subway entrance and exit as follows Figure 4 As shown, find the minimum cut set of the event tree, that is, the minimum combination of events that lead to the occurrence of the upper-level event. The meaning of each event is as follows Figure 7 As shown, there are 12 minimum cut sets through calculation. The specific formula is as follows:

[0187] T=A1+A2+A3=X1X2+X3X4X5+X3X4X6+X3X4X7+X3X4X8+X3X4X9+X3X4X 10 +X3X4X 11 +X 12 +X 13 +X 14 +X 15

[0188] Then, the structural importance of each lower-level event is calculated, and the result reflects the relative importance of each lower-level event in affecting the upper-level event.

[0189]

[0190] The structural importance ranking is: X3=X4>X 12 =X 13 =X 14 =X 15 >X1=X2>X5=X6=X7=X8=X9=X 10 =X 11

[0191] Through the structural importance analysis, it can be found that the top structural importance rankings are: extreme rainstorm (X3), urban waterlogging (X4), subway failure (X 12 ), emergency management chaos (X 13 ), the emergency plan is not perfect (X 14 ), the accuracy of rainstorm warning is low (X 15), inappropriate self-rescue behavior (X1), and low rescue efficiency (X2). These lower-level events have a significant impact on subway flood disasters. Therefore, by analyzing the structural importance of each lower-level event, we can effectively identify which events play a key role in the occurrence of disasters. It not only helps us to clarify the emergency events of the accident, but also provides a scientific basis for taking measures to improve the safety of the entire subway system. By focusing on events with high structural importance, flood control equipment can strengthen the safety of weak links in outdoor areas;

[0192] Then, construct the accident tree of subway entrance and exit, such as Figure 5 As shown in the figure, by conducting a detailed analysis of the causes of disasters, the frequency and probability of each lower-level event are statistically calculated. This process not only helps to understand the laws of disaster occurrence, but also quantifies the impact of each lower-level event on the disaster, thereby providing a basis for formulating disaster prevention and mitigation measures; through statistical analysis of frequency and probability, the potential risk of disasters can be more accurately assessed, and high-frequency, high-risk lower-level events can be identified, which not only provides a quantitative scientific basis for disaster management, but also helps flood control equipment to more effectively optimize resource allocation and concentrate on solving the most critical safety hazards. The frequency and probability of each lower-level event are shown in the following table:

[0193]

[0194] Therefore, according to the upper probability formula, the probability of the upper event subway flood disaster is 0.103;

[0195] In this embodiment, the Bayesian network can perform forward prediction and backward diagnosis, thereby quantitatively analyzing the accident. When the risk result is determined, according to the reverse reasoning algorithm of the Bayesian network, the probability of subway flood disaster is determined to be 100%, and the direction of the arrows between the nodes in the network is inverted. The Bayesian formula is as follows:

[0196]

[0197] Among them, P(X i ) is the underlying event X i The prior probability, P(X i |T) is the lower layer event X when a subway flood disaster T occurs i The probability of occurrence, that is, X i The posterior probability of P(T) is the probability of subway flood disaster;

[0198] According to the probability of the subsequent occurrence of the lower-level events and the degree of impact on the upper-level events, the importance of each lower-level event can be comprehensively evaluated and obtained. This analysis method not only considers the possibility of the occurrence of lower-level events, but also considers their role and contribution in the entire system failure or accident chain. Through comprehensive evaluation, it can be clear which lower-level events have a critical role in triggering upper-level events, thereby providing a more accurate reference for disaster and flood control equipment safety management and prevention. Determining the importance of lower-level events is conducive to giving priority to the most threatening factors in risk management, optimizing protective measures, enhancing system reliability and improving risk resistance, and then adopting effective emergency strategies. The specific formula for establishing key importance is as follows:

[0199]

[0200] Among them, P(T occurs|X i P(Toccurrence|X i does not occur) indicates the probability of the occurrence of the upper-level event under the condition that the lower-level event does not occur; the critical importance can integrate the prior probability and posterior probability of the lower-level event, can eliminate the influence of the one-sided analysis of the disaster-causing factors, and has a strong comprehensiveness; through this method, flood control equipment can clearly grasp the complexity and diversity of disaster factors, so as to formulate more accurate prevention and control measures. This highly comprehensive analysis method greatly improves the depth and accuracy of disaster risk assessment, and helps to comprehensively improve the safety and stability of the system; the specific conditional probability of each lower-level event and the critical importance of the lower-level event are shown in the following table:

[0201]

[0202] It can be seen that the most influential factors in subway flood disasters are: backflow of water in the station (X5), low height of subway entrances and exits (X6), paralysis of the drainage system in the station (X7), and failure of the retaining wall (X8). 11 ); medium-impact disaster factors include: low rescue efficiency (X2), extreme rainstorms (X3), urban waterlogging (X4), poor anti-leakage performance in stations (X8), insufficient flood control materials (X9), 10 ), emergency management chaos (X 13 ), the emergency plan is not perfect (X 14 ), low accuracy of rainstorm warning (X 15 ); the disaster-causing factors with less impact are: inappropriate self-rescue behavior (X1), power outage or leakage in the station (X9), subway failure (X 12 ). When one or more of the key disaster-causing factors occur, the possibility of subway flood disasters is relatively high, and subway flood control equipment needs to respond quickly to these key disaster-causing factors.

[0203] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0204] A person skilled in the art may understand that all or part of the steps of the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0205] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0206] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0207] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A rapid response method for flood control equipment at subway entrances and exits based on water level situation analysis, characterized in that: The method comprises the following steps: Step S1: Collect and obtain surveillance video images of different subway entrances and exits, read the video images frame by frame, and obtain a water level image sequence at different times; Step S2: applying the inverse perspective transformation method of point-to-point control to correct the distorted image; Step S3: performing target area extraction and threshold segmentation operations on the corrected water level image; Step S4: extract the water level position of each frame from the processed image sequence, and analyze the water level situation according to the position change of the water level; Step S5: Generate a flood control decision tree model based on the water level situation, which is divided into different risk levels, and the flood control equipment takes corresponding flood control measures according to the level.

2. The rapid response method for subway entrance and exit flood control equipment based on water level situation analysis according to claim 1 is characterized by: The step S1 specifically includes: Step S101: Acquire surveillance video images of multiple subway entrances and exits; Step S102: reading the collected surveillance videos of subway entrances and exits frame by frame using a dynamic synchronous frame extraction method; Step S103: Divide the read video image into water level image sequences at different moments according to time.

3. According to claim 1, a rapid response method for subway entrance and exit flood control equipment based on water level situation analysis is characterized by: The step S2 comprises: Step S201: randomly determine four points in the three-dimensional space to record the position coordinates of four feature points in the water level image; Step S202: Based on the position coordinates determined in S201, specify the real position coordinates under the normal viewing angle, and calculate the corresponding relationship between the two sets of position coordinates; Step S203: converting the monitoring imaging coordinate system into an abstract pixel coordinate system; The specific method of converting the pixel coordinate system in step S203 is as follows: First, select the four pixels at the outermost edge of the image as feature points, and record the coordinates of these four control points, denoted as (x1, y1), (x2, y2), (x3, y3), (x4, y4), and use their corresponding actual coordinates (u1, v1), (u2, v2), (u3, v3), (u4, v4) as known quantities; use perspective transformation to transform the target image into another plane image, and the transformation process is calculated using the following formula: Among them, (x, y) is the pixel coordinates of the perspective distorted image, (u, v) is the pixel coordinates of the corrected image, a, b, c, d, e, f, p, q are perspective transformation parameters. For the convenience of calculation, the mathematical formula needs to be converted into a matrix form to obtain the corresponding perspective transformation matrix. The rewritten matrix form is as follows: Then, the coordinates of the four sets of feature points are substituted into the matrix formula to obtain: Let this matrix be: UV = A × M; According to the above formula, the perspective transformation matrix can be obtained as: M=A -1 UV; Finally, the obtained perspective transformation matrix M is used to process each pixel in the target image to obtain the corrected water level image.

4. The method for rapid response of subway entrance and exit flood control equipment based on water level situation analysis according to claim 1 is characterized by: The step S3 comprises: Step S301: manually calibrate the coordinates of the four corner points of the backboard above the water surface based on the corrected water level image. The trapezoidal area enclosed by them is the backboard part of the water level gauge above the water surface. The two waistlines of the trapezoidal area are respectively extended longitudinally to the bottom of the image to obtain the complete backboard area; Step S302: Based on the complete backplane area obtained in step S301, the image in RGB format is converted into HSV color space to obtain an image in HSV format; Step S303: performing color threshold segmentation and positioning on the water level image in HSV format; The specific method of converting the RGB format image to the HSV format in step S302 is as follows: First, normalize the RGB values ​​to the range of 0 to 1 by dividing the value of each color by 255. The specific formula is as follows: R n =R / 255,G n =G / 255,B n =B / 255; Find the maximum value C among the three normalized values max =max(R n ,G n ,B n ) and minimum value C min =min(R n ,G n ,B n ), according to C max and C min Calculate the hue H, saturation S, and brightness V. The calculation method of H, S, and V depends on which color channel contains the maximum value. The specific calculation method is as follows: V=C max ; 5. A rapid response method for subway entrance and exit flood control equipment based on water level situation analysis according to claim 4, characterized in that: In step S303, the specific steps of color threshold segmentation and positioning of the water level image are as follows: first, for the features in the water level image that are different from other background colors, the image after the target area is extracted is threshold segmented based on red to obtain a binary image M(x, y); the red range is fuzzy processed as: H[0-10&156-180], S[43-255], V[46-255], by setting the HSV three channels, the image is segmented according to the threshold of the H channel to obtain a binary image containing all red spot areas, and the threshold segmentation method for the red area is as follows: Then, the m×n neighborhood average calculation method is used to calculate the mean of the binary image M(x,y) to reduce the small red part in the background and obtain the final binary image. Finally, for the image Perform horizontal and vertical projections to determine the position of the water level line in the image.

6. The rapid response method for subway entrance and exit flood control equipment based on water level situation analysis according to claim 1 is characterized by: The step S4 comprises: Step S401: extracting target light spots with changing image background from the water level image processed in S3, calculating and identifying them using a target tracking algorithm based on adaptive mean shift enhancement, extracting the number, size, shape, and movement information of the light spots in the image sequence, and obtaining a time series of water levels at different subway entrances and exits; Step S402: Check the stability of the water level data at the subway inlet and outlet, perform seasonal difference processing on the unstable monthly water level data, and design a seasonal change periodicity prediction model. The model consists of seasonal autoregressive values, seasonal difference values, and error terms. The autoregressive part introduces seasonal parameters to obtain seasonal autoregressive values: Among them, σ(A) describes the relationship between the water level data and its own hysteresis value; σ i is the coefficient corresponding to the autoregressive order n, A i is an operator that lags by i time units and is used to represent time lag; σ i *A i It represents the relationship between water level data and its data lagged by i time units; Introducing seasonal parameters into the difference part can obtain seasonal difference values ​​(1-A s ) d : Among them, (1-A) d It is used to convert non-stationary time series into stationary time series. d represents the difference order, that is, the number of times the water level data is differentiated; (1-A s ) d It is used to convert non-stationary time series with seasonal characteristics into stationary time series. d is the seasonal difference order, that is, the number of times the water level data is seasonally differenciated; s is the seasonal parameter; The established seasonal change periodicity prediction model Ω is: Ω=σ(A)φ(A s (1-A) d (1-A s ) d y i =c+σ(A)*α i ; Among them, c is a constant term, which is used to represent the average trend of water level; α i is the error term, which is used to represent the random fluctuation of water level; Step S403: Establish a seasonal cycle prediction model Ψ, and calculate the error sum of squares e of the seasonal change periodicity prediction model Ω and the seasonal cycle prediction model Ψ respectively. j , based on the principle of minimizing the overall error sum of squares, the weights of each prediction model are assigned, and the combined model for subway entrance and exit water level prediction is obtained; the combined weight of the jth model is: Step S404: input the historical subway entrance and exit water level data into the established combined model, and obtain future water level data according to seasonal attributes for subsequent flood control equipment analysis.

7. The rapid response method for subway entrance and exit flood control equipment based on water level situation analysis according to claim 4 is characterized by: The target tracking algorithm based on adaptive mean shift enhancement in step S401 specifically describes the histogram of the detection area respectively, and performs iterative calculations repeatedly, traversing the entire image through the process of moving convergence. When the calculation converges to a moving interval less than a set threshold, the recognition of the detection target is completed, and the result of this time is passed to the next frame image as the result of the search, and such a cycle is repeated until the function of the image to be detected converges.

8. The rapid response method for subway entrance and exit flood control equipment based on water level situation analysis according to claim 6 is characterized by: The Ψ model established in step S403 consists of three parts: a season reset gate, a season update gate and a season hidden state. The specific method is as follows: First, the Season Update Gate Z t It is determined by the state value of the previous moment and the input value of the current moment. When the value of the seasonal update gate is larger, it means that the current input neuron needs to retain more information, and the information of the previous moment is retained less. The specific expression is as follows: z t =σ(w xz *x t +w hz *h t-1 +b z ); Among them, x t is the input sequence, h t is the state of the hidden layer, h t-1 is the state of the hidden layer in the last second, w xz is the weight matrix from the input layer to the seasonal update gate, w hz represents the weight matrix of the recurrent connection, σ(·) represents the Sigmoid function, b z Update the gate bias for the seasons, which is used to adapt to different input data by adjusting the threshold of the activation function. It also improves the gradient flow in back propagation, helps avoid network symmetry, and allows each neuron to learn unique features, thereby improving the network's representation ability and training effect; Then, the season resets the gate t It is determined by the state value at the previous moment and the input gate at the current moment. When the value of the reset gate is larger, it means that more information is retained at the previous moment. The specific calculation formula is as follows: r t =σ(w xr *x t +w hr *h t-1 +b r ); Among them, w xr is the weight matrix of the input layer to the seasonal reset gate, w hr represents the weight matrix of the recurrent connection, σ(·) represents the Sigmoid function, b r Indicates the bias of the seasonal reset gate; Finally, the output o of the memory unit can be obtained t for: Among them, c t is the gating signal, h t is the hidden layer state, w xc is the weight matrix from the input layer to the candidate state, w hc represents the weight matrix of the recurrent connection, represents element-wise multiplication, Represents the Tanh function.

9. The method for rapid response of subway entrance and exit flood control equipment based on water level situation analysis according to claim 1 is characterized by: The step S5 comprises: Step S501: define a risk quantification formula based on the future water level data predicted in S4 to quantitatively assess the risk of flooding. The formula is as follows: R(E i )=B*P(E i ) α *(I(E i )*V(E i )) β ; Among them, E i is the flood event to be calculated; P(E i ) is event E i Probability of occurrence; I(E i ) is event E i The degree of impact when it occurs; V(E i ) is event E i The vulnerability of the model to the event; B is the risk adjustment factor, which can be specifically expressed as: B = 1 + ε, where ε is the risk preference coefficient; α and β are weight factors for the probability and degree of impact, which are obtained by studying the probability and degree of impact of flood events at subway entrances and exits; Step S502: define four risk levels of low risk, medium risk, high risk and extremely high risk, and set a two-way Bayesian decision model for each entrance and exit to analyze the cause of the accident and the possible harm, so that the model can generate preventive measures before the accident and control measures after the accident.

10. The method for rapid response of subway entrance and exit flood control equipment based on water level situation analysis according to claim 4, characterized in that: The two-way Bayesian decision model established in step S502 comprises the following steps: First, construct an accident tree. The specific method is as follows: determine the upper-level events, intermediate events, and lower-level events, and use logic gates to connect the upper and lower-level events. If all lower-level events must occur at the same time to cause the upper-level event to occur, use an AND gate to connect; conversely, if only one or some of the lower-level events can cause the upper-level event to occur, use an OR gate to connect; Then, the event tree is constructed, specifically determining the initial event, determining the safety function and drawing the event tree; Furthermore, the structural importance in the bidirectional Bayesian decision model is calculated. By calculating the number of times each lower-level event appears in the minimum cut set, the structural importance of each lower-level event is determined, and the minimum cut set of the fault tree is obtained, which is the minimum combination of events that lead to the occurrence of the upper-level event. Structural importance analysis can evaluate the reliability and safety of the model and help determine which safety measures to prioritize to improve the safety of the model. The calculation formula of structural importance is as follows: Among them, q represents the total number of minimum cut sets, r i represents the i-th minimum cut set, t i It is represented as the number of lower-level events of the i-th minimum cut set, and n represents the number of minimum cut sets; Finally, the probability of the upper-level events of the model is calculated so that the flood control equipment can respond quickly. The specific method is: replace all events in the original event tree with events in the minimum cut set to obtain a new equivalent tree. In the equivalent tree, the upper-level events are equal to the union of the minimum cut sets, that is, all events in the minimum cut sets must occur at the same time to cause the occurrence of the upper-level events. It is known that there are r minimum cut sets in the accident tree: E1, E2, ..., E k 、E r , then the upper-level event is expressed as: The probability of the upper event is expressed as: Among them, p i represents the probability of the occurrence of the lower-level event i, k, l, and r represent the numbers of the minimal cut sets, and the size relationship of the numbers is k < l < r; i ≤ k ≤ l ≤ r represents the combination order of the kth and lth cut sets in the minimal cut set, x i ∈ E k represents the ith lower-level event belonging to the kth minimal cut set.

Images