Digitization method of urban safety risk assessment standard

Abstract

The invention discloses a digitalization method of an urban safety risk assessment standard, which comprises the steps of identifying risk factors for historical accident time sequence data, performing correlation analysis to obtain a correlation function, further constructing an urban network topology structure, determining a prior probability of a node according to the similarity of the node and pre-disaster data, and determining the probability of the node according to the prior probability of the node. Generating a simulated accident scene, adding a disturbance item to simulate an emergency risk, combining with the prior probability of the nodes to obtain a risk propagation path, constructing a risk assessment model, assessing the space-time influence range of the nodes on the propagation path, and performing weighted summation on the assessment results of the nodes to obtain an urban safety risk assessment result. According to the method, a network topology structure is constructed by multi-stage analysis of historical accident data, so that risk propagation simulation is more accurate, urban risk assessment can be accurately quantified, and meanwhile, the method has relatively good interpretability.

Description

Technical Field

[0001] The invention relates to the field of public safety, and in particular to a digital method for evaluating urban safety risk standards. Background Art

[0002] With the rapid development of urbanization, the scale of cities continues to expand, the population is highly concentrated, and the functions are complex and diverse. Numerous buildings, transportation facilities, energy supply and public service facilities and other key elements are intertwined, making cities face many security risks such as natural disasters, accidents, public health incidents and social security incidents.

[0003] In the process of urban safety management, it is crucial to accurately assess urban safety risks and formulate corresponding strategies. However, traditional urban safety risk assessment methods have many limitations. They mostly rely on manual field surveys, paper records and simple empirical judgments, with a single data source, and it is difficult to fully cover the complex information of various elements in the city.

[0004] A digital method for urban safety risk assessment standards integrates multi-source big data and uses advanced algorithm models to make up for the shortcomings of traditional assessment methods, thereby providing a more accurate, real-time, comprehensive and forward-looking risk assessment solution for urban safety management, effectively ensuring the safe, stable operation and sustainable development of the city. Summary of the invention

[0005] The purpose of the present invention is to provide a digital method for urban safety risk assessment standards.

[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0007] The first aspect of the present invention provides a digital method for urban safety risk assessment standards, comprising:

[0008] S101 identifies risk factors on historical accident time series data to obtain a risk factor set; the historical accident time series data includes pre-disaster data, mid-disaster data, and post-disaster data;

[0009] S102 performs correlation analysis on the risk factor set to obtain a risk factor correlation function;

[0010] S103 constructing a city network topology structure based on the risk factor correlation function, and determining a priori probability of each node according to the similarity between each node and the pre-disaster data;

[0011] S104 generates a simulated accident scenario, adds a disturbance item to simulate a sudden risk event, and obtains a propagation path of the risk in the urban network topology structure in combination with the prior probability of each node;

[0012] S105 obtains the spatiotemporal influence range of each node traversed by the propagation path, constructs a risk assessment model to assess the spatiotemporal influence range of each node, and performs weighted summation on the assessment results of each node to obtain a city safety risk assessment result.

[0013] As a further method, the method of identifying risk factors for historical accident time series data to obtain a risk factor set includes:

[0014] Extract environmental features from pre-disaster data, extract disaster characteristics from disaster data, and extract loss degree characteristics from post-disaster data, and combine them as a set of urban safety risk features;

[0015] Calculate the degree of correlation between features, the expression is:

[0016]

[0017] Among them, X i and Y j are the i-th item of feature sequence X and the j-th item of feature sequence Y, respectively. m and n are the number of feature sequences X and Y, respectively. α ij and β ij X i and Y j The linear correlation weight and nonlinear correlation weight between ij For X i The positive covariance power adjustment coefficient, μ ij Y j The positive covariance power adjustment coefficient, δ ij Y j The absolute value of the covariance power adjustment coefficient, Cov(X i ,Y j ) is X i and Y j The covariance between and X i and Y j The norm of ;

[0018] If the correlation between environmental characteristics and disaster characteristics, and between environmental characteristics and loss degree characteristics are greater than the preset threshold, the corresponding environmental characteristics will be used as risk factors to obtain a risk factor set.

[0019] As a further method, the method of performing correlation analysis on the risk factor set to obtain a risk factor correlation function includes:

[0020] A support vector machine model was constructed, and the risk factors for correlation analysis were used as input and output to train the model. The kernel function of the model was used to represent the correlation function, and the expression was:

[0021]

[0022] Where N is the number of support vectors, α i is the positive weight of the i-th support vector on the prediction result, is the negative weight of the i-th support vector on the prediction result, γ is the width of the kernel function, x is the current risk factor input vector, x i is the risk factor input vector of the i-th support vector, and b is the bias term.

[0023] As a further method, the method for constructing a city network topology structure based on the risk factor association function includes:

[0024] Each functional area in the city is defined as a node. The node types include geographical area nodes, infrastructure nodes, and personnel gathering nodes. Each node is assigned an attribute value based on the risk element data.

[0025] The function of determining the edge between nodes based on the multilayer perceptron neural network is as follows: the number of neurons in the input layer and the output layer are respectively set to the number of risk factors of the two nodes, and the correlation function between the risk factors is used as the mapping function of the corresponding node in the hidden layer of the neural network to obtain the edge function;

[0026] Establish a data collection system to continuously collect risk element data, update node attribute values ​​and edge functions based on the determined risk element data and risk element association functions, and dynamically adjust the network topology.

[0027] As a further method, the method of determining the prior probability of each node according to the similarity between each node and the pre-disaster data includes:

[0028] Get the pre-disaster data corresponding to the node and calculate the similarity between the node and the pre-disaster data. The expression is:

[0029]

[0030] Among them, s i is the similarity between the i-th node and the pre-disaster data, k is the number of features, ω j is the weight of the jth feature, x j and ij are the jth eigenvalue of the node and the jth eigenvalue of the i-th sample in the historical pre-disaster data, max(ξ j ) and min(ξ j ) are the maximum and minimum values ​​of the jth feature in the historical pre-disaster data samples, ∈ is a smoothing term, which is 0.01;

[0031] Calculate the prior probability of each node, the expression is:

[0032]

[0033] Among them, P(s i ) is the prior probability of the ith node, e is a constant, μ is the mean of the similarity distribution, σ is the standard deviation of the similarity, α1 is the amplitude adjustment parameter of the sine function, β1 is the period adjustment parameter of the sine function, γ1 is the translation parameter of the sine function, n is the total number of nodes, s j is the similarity between the jth node and the pre-disaster data.

[0034] As a further method, the method of generating a simulated accident scenario and adding a disturbance item to simulate an unexpected risk event includes:

[0035] A deep convolutional generative adversarial network is selected to construct a generator network and a discriminator network, which are trained based on historical accident case data. The latent space distribution of historical accident data is input into the generator network to obtain a set of simulated accident scenarios.

[0036] Add disturbance terms to simulate accident scenarios. The disturbance term expression is:

[0037]

[0038] Among them, t is the time variable, m is the total number of disturbance terms, P i is the weight of the ith disturbance term, χ i is the fluctuation adjustment coefficient of the ith disturbance term, ψ i is the fluctuation frequency parameter of the ith disturbance term, t 0i is the starting time of the ith disturbance term, ω i is the Gaussian attenuation coefficient of the ith disturbance term, H(·) is a step function, when t≥t 0i When H(tt 0i )=1, indicating that the disturbance term begins to take effect, otherwise the disturbance term has no effect.

[0039] As a further method, the method of obtaining the propagation path of the risk in the urban network topology structure in combination with the prior probability of each node includes:

[0040] The nodes in the urban topology network are used as nodes of the Bayesian network, and the conditional probability table is determined based on the prior probability of the nodes to convert the simulated accident scenarios into constraint conditions;

[0041] The probability of a node propagating to an adjacent node is determined based on the maximum a posteriori probability estimation method under constraints. The expression is:

[0042]

[0043] in, For node N i In new state Under the condition that node N j Transition to a new state The probability of For node N i In new state Under the condition that node N j Transition to a new state The conditional probability of For node N i In new state The prior probability of α is the weight parameter of the prior probability, C is the constraint factor, β is the weight factor of the constraint, and D ij For node N i With node N j The distance, S j For node N j All possible state values ​​of For node N i In new state Under the condition that node N j The probability of being in various possible states;

[0044] For each simulated accident scenario, the paths with the highest probability of propagating to adjacent nodes are selected from the nodes and combined to obtain the risk propagation path in the urban network topology.

[0045] As a further method, the method of constructing a risk assessment model to assess the spatiotemporal impact range of each node and performing weighted summation on the assessment results of each node to obtain the urban safety risk assessment result includes:

[0046] A risk assessment model was constructed based on a three-dimensional convolutional neural network. The historical recovery time, loss degree, and the spatial and temporal impact range of historical accidents were used as training sets to train the model. The predicted recovery time and loss degree were used as evaluation indicators.

[0047] Calculate the importance of each node, the expression is:

[0048]

[0049] Among them, I i is the importance of node i, K i is the degree of node i in the topological network structure, q i is the resource flow of node i, C i is the expected recovery time of node i, η is the risk resistance sensitivity coefficient, R iis the risk resistance assessment value of node i, μ R and σ R are the mean and standard deviation of the risk resistance assessment values ​​of all nodes, θ is the node association strength weight coefficient, S i is the mean of the association degree between node i and the rest of the nodes;

[0050] The evaluation index of each node is weighted and summed based on the importance of the node to obtain the risk assessment value of the entire urban topological network structure.

[0051] In a second aspect, an embodiment of the present application further provides an electronic device, comprising: a processor; and a memory arranged to store computer executable instructions, wherein when the executable instructions are executed, the processor executes the method steps described in the first aspect.

[0052] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores one or more programs. When the one or more programs are executed by an electronic device including multiple applications, the electronic device executes the method steps described in the first aspect.

[0053] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0054] (1) By analyzing the time series data of historical accidents before, during and after disasters, the present invention can comprehensively capture risk factors, so that risk assessment is no longer limited to a single stage, but covers the entire life cycle of risks, thereby more systematically identifying and understanding urban safety risks.

[0055] (2) By adding disturbance items to simulate sudden risk events, the present invention can more realistically reflect the uncertainty and suddenness of risks in reality, make the acquisition of risk propagation paths more in line with actual conditions, and enhance the reliability of the assessment results.

[0056] (3) The present invention obtains the city safety risk assessment result by weighted summing up the assessment results of each node based on the node importance, so that the final assessment result can more accurately reflect the overall safety risk of the city. At the same time, based on the assessment result of a single node, the safety of key areas and key links can be prioritized under limited resource conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 The present invention is a flowchart of the steps of a digital method for evaluating urban safety risk standards in an embodiment of the present invention.

[0058] Figure 2 It is a schematic diagram of the structure of an electronic device in an embodiment of this specification. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0060] Reference Figure 1 As shown, the present invention provides a digital method for urban safety risk assessment standards, including:

[0061] S101 identifies risk factors on historical accident time series data to obtain a risk factor set; the historical accident time series data includes pre-disaster data, mid-disaster data, and post-disaster data;

[0062] In the actual evaluation, traffic data of a large city in the past three years was collected:

[0063] Pre-disaster data included road infrastructure data, with a total road mileage of about 5,000 kilometers, including 1,200 kilometers of main roads, 1,700 kilometers of secondary roads, and 2,100 kilometers of branch roads; road width data, with an average width of 40 meters for main roads, 25 meters for secondary roads, and 15 meters for branch roads; a total of 8,700 traffic lights distributed on different road types; and vehicle ownership data, about 2 million vehicles, and growing at an annual rate of 8%;

[0064] The disaster data covers traffic flow data of various road sections at different time periods (7-9 am during the morning peak and 17-19 pm during the evening peak). The average traffic flow of main roads during the morning peak is 820 vehicles per lane per hour, 410 vehicles per secondary road, and 200 vehicles per lane per hour on branch roads; the average traffic flow of main roads during the evening peak is 1,100 vehicles per lane per hour, 450 vehicles per secondary road, and 300 vehicles per lane per hour on branch roads.

[0065] Post-disaster data include economic losses, time delays, and casualties caused by traffic accidents;

[0066] In the actual assessment, environmental characteristics are extracted from pre-disaster data, disaster characteristics are extracted from mid-disaster data, and loss degree characteristics are extracted from post-disaster data. The degree of correlation between the characteristics is calculated, and the preset threshold of the degree of correlation is set to 0.5. Factors that meet the threshold conditions are identified as a set of risk factors, including: road width, motor vehicle ownership and public transportation capacity ratio.

[0067] S102 performs correlation analysis on the risk factor set to obtain a risk factor correlation function;

[0068] It needs to be explained that through correlation analysis, the complex interactions among risk factors can be obtained, which helps to reasonably set the relationship between nodes and edges when constructing the topological structure of the urban network;

[0069] In the actual assessment, the risk factors of road width, motor vehicle ownership and public transport capacity ratio were used as input and output training models respectively to obtain the risk factor association function.

[0070] S103 constructing a city network topology structure based on the risk factor correlation function, and determining a priori probability of each node according to the similarity between each node and the pre-disaster data;

[0071] It should be explained that constructing the urban network topology based on the risk factor correlation function is to regard the city as a complex network system and abstract each functional area as a node. The topological structure constructed in this way can intuitively present the potential path and logical structure of risk propagation; and the prior probability is determined according to the similarity between each node and the pre-disaster data, because the pre-disaster data reflects many characteristics of the city under normal conditions. High similarity means that the node is closer to the relatively safe state in the past under the current state, and the prior probability of a risk event is relatively low; otherwise, it is higher. The prior probability provides an initial possibility judgment basis for the subsequent calculation of the risk propagation path under the simulated accident scenario, so that the risk assessment process can comprehensively consider the historical normality and the current node characteristics, and improve the accuracy and scientificity of the assessment.

[0072] In the actual assessment, the city is divided into multiple traffic area nodes, including city center nodes, railway station perimeter nodes, residential area nodes, school area nodes, commercial area nodes and industrial area nodes. The function of the edges between nodes is determined based on the multi-layer perceptron neural network to obtain the urban network topology. Through traffic monitoring cameras, geomagnetic sensors and other data collection systems, risk element data is continuously collected to monitor changes in traffic volume and road construction conditions in real time. The node attribute values ​​and edge functions are updated according to these data and risk element association functions to dynamically adjust the network topology. The pre-disaster data corresponding to the node are obtained, and the similarity between the node and the pre-disaster data is calculated. For the school area node, there are four characteristics: road grade, road width, traffic light density, and vehicle ownership. The calculated similarity between the school area node and the pre-disaster data is 0.7, and the prior probability of the school area node is 0.42.

[0073] S104 generates a simulated accident scenario, adds a disturbance item to simulate a sudden risk event, and obtains a propagation path of the risk in the urban network topology structure in combination with the prior probability of each node;

[0074] It needs to be explained that by constructing a variety of simulation scenarios, we can cover accident situations of different types and severity, such as traffic accidents, traffic paralysis caused by bad weather, etc., so as to comprehensively study the manifestation of risks in urban systems; adding disturbance items to simulate sudden risk events takes into account that in actual situations, risks do not exist in isolation, and there are often various sudden factors involved, such as sudden changes in wind direction in fire scenarios, unexpected road collapses in traffic scenarios, etc. These disturbance items can make the simulation closer to the real and complex and changeable real environment; combining the prior probability of each node to obtain the risk propagation path, because the prior probability reflects the initial risk tendency of the node. Under the influence of simulation scenarios and disturbance items, based on the prior probability, using methods such as Bayesian networks, we can infer the possibility of risk propagation from the accident node to the adjacent nodes, and then determine the most likely propagation path, which provides a key basis for timely formulation of targeted risk prevention and control and response strategies;

[0075] In the actual evaluation, a deep convolutional generative adversarial network was used to construct the generator network and the discriminator network. The training was based on historical traffic congestion accident case data. The latent spatial distribution of historical traffic data was input into the generator network to obtain a set of simulated traffic congestion accident scenarios. One of the scenarios was a traffic accident on a main road in the downtown business district, which caused the traffic capacity of the road section to drop by 60%. Disturbance items were added to simulate the accident scenario. The first disturbance item was that the road was slippery due to bad weather, and the vehicle speed was reduced by 30%. The second disturbance item was temporary construction of nearby roads. The third disturbance item was the instantaneous increase in surrounding traffic volume caused by the end of a large-scale event.

[0076] In the actual assessment, for the case of risk transmission from downtown nodes to adjacent residential nodes, the probability of node transmission to adjacent nodes is determined based on the maximum a posteriori probability estimation method under constraints. Under the influence of severe weather disturbance terms, the probability of transmission from downtown business district nodes to adjacent residential nodes is calculated to be 0.72. For each simulated accident scenario, the paths with the highest probability of transmission to adjacent nodes are selected from the nodes and combined to obtain the risk transmission path in the urban network topology.

[0077] S105 obtains the spatiotemporal influence range of each node traversed by the propagation path, constructs a risk assessment model to assess the spatiotemporal influence range of each node, and performs weighted summation on the assessment results of each node to obtain a city safety risk assessment result.

[0078] In the actual assessment, a risk assessment model was constructed based on a three-dimensional convolutional neural network. The historical traffic recovery time (the time it took for traffic in various regions to return to normal in past congestion accidents, averaging 30 minutes to 2 hours), loss level (economic loss, vehicle delay time) and historical accident spatiotemporal impact range (road length and regional area of ​​congestion spread) were used as the training set to train the model, and the predicted recovery time and loss level were used as evaluation indicators. For the nodes in the city center commercial district, its degree in the topological network structure is 6, the resource flow is 100, the risk resistance capacity assessment value is 0.6, the risk resistance capacity sensitivity coefficient is set to 0.2, the mean of the risk resistance capacity assessment values ​​of all nodes is 0.55, the standard deviation is 0.1, the node association strength weight coefficient is 0.3, and the mean of the degree of association between the node and the remaining nodes is 0.45. The importance of the nodes in the city center commercial district is calculated to be 0.78. After weighted summation of the evaluation indicators of all nodes, the traffic congestion safety risk assessment value of the entire city is 0.65.

[0079] In this embodiment, the method of identifying risk factors for historical accident time series data to obtain a risk factor set includes:

[0080] Extract environmental features from pre-disaster data, extract disaster characteristics from disaster data, and extract loss degree characteristics from post-disaster data, and combine them as a set of urban safety risk features;

[0081] Calculate the degree of correlation between features, the expression is:

[0082]

[0083] Among them, X i and Y j are the i-th item of feature sequence X and the j-th item of feature sequence Y, respectively. m and n are the number of feature sequences X and Y, respectively. α ij and β ij X i and Y j The linear correlation weight and nonlinear correlation weight between ij For X i The positive covariance power adjustment coefficient, μ ij Y j The positive covariance power adjustment coefficient, δ ij Y j The absolute value of the covariance power adjustment coefficient, Cov(X i ,Y j ) is X i and Y j The covariance between and X i and Yj The norm of ;

[0084] If the correlation between environmental characteristics and disaster characteristics, and between environmental characteristics and loss degree characteristics are greater than the preset threshold, the corresponding environmental characteristics will be used as risk factors to obtain a risk factor set.

[0085] In this embodiment, the method of performing correlation analysis on the risk factor set to obtain a risk factor correlation function includes:

[0086] A support vector machine model was constructed, and the risk factors for correlation analysis were used as input and output to train the model. The kernel function of the model was used to represent the correlation function, and the expression was:

[0087]

[0088] Where N is the number of support vectors, α i is the positive weight of the i-th support vector on the prediction result, is the negative weight of the i-th support vector on the prediction result, γ is the width of the kernel function, x is the current risk factor input vector, x i is the risk factor input vector of the i-th support vector, and b is the bias term.

[0089] In this embodiment, the method for constructing a city network topology structure based on the risk factor association function includes:

[0090] Each functional area in the city is defined as a node. The node types include geographical area nodes, infrastructure nodes, and personnel gathering nodes. Each node is assigned an attribute value based on the risk element data.

[0091] The function of determining the edge between nodes based on the multilayer perceptron neural network is as follows: the number of neurons in the input layer and the output layer are respectively set to the number of risk factors of the two nodes, and the correlation function between the risk factors is used as the mapping function of the corresponding node in the hidden layer of the neural network to obtain the edge function;

[0092] Establish a data collection system to continuously collect risk element data, update node attribute values ​​and edge functions based on the determined risk element data and risk element association functions, and dynamically adjust the network topology.

[0093] In this embodiment, the method of determining the prior probability of each node according to the similarity between each node and the pre-disaster data includes:

[0094] Get the pre-disaster data corresponding to the node and calculate the similarity between the node and the pre-disaster data. The expression is:

[0095]

[0096] Among them, s i is the similarity between the i-th node and the pre-disaster data, k is the number of features, ω j is the weight of the jth feature, x j and ij are the jth eigenvalue of the node and the jth eigenvalue of the i-th sample in the historical pre-disaster data, max(ξ j ) and min(ξ j ) are the maximum and minimum values ​​of the jth feature in the historical pre-disaster data samples, ∈ is a smoothing term, which is 0.01;

[0097] Calculate the prior probability of each node, the expression is:

[0098]

[0099] Among them, P(s i ) is the prior probability of the ith node, e is a constant, μ is the mean of the similarity distribution, σ is the standard deviation of the similarity, α1 is the amplitude adjustment parameter of the sine function, β1 is the period adjustment parameter of the sine function, γ1 is the translation parameter of the sine function, n is the total number of nodes, s j is the similarity between the jth node and the pre-disaster data.

[0100] In this embodiment, the method of generating a simulated accident scenario and adding a disturbance item to simulate an unexpected risk event includes:

[0101] A deep convolutional generative adversarial network is selected to construct a generator network and a discriminator network, which are trained based on historical accident case data. The latent space distribution of historical accident data is input into the generator network to obtain a set of simulated accident scenarios.

[0102] Add disturbance terms to simulate accident scenarios. The disturbance term expression is:

[0103]

[0104] Among them, t is the time variable, m is the total number of disturbance terms, P i is the weight of the ith disturbance term, χ i is the fluctuation adjustment coefficient of the ith disturbance term, ψ i is the fluctuation frequency parameter of the ith disturbance term, t 0i is the starting time of the ith disturbance term, ω i is the Gaussian attenuation coefficient of the ith disturbance term, H(·) is a step function, when t≥t 0i When H(tt 0i )=1, indicating that the disturbance term begins to take effect, otherwise the disturbance term has no effect.

[0105] In this embodiment, the method of obtaining the propagation path of risk in the urban network topology structure in combination with the prior probability of each node includes:

[0106] The nodes in the urban topology network are used as nodes of the Bayesian network, and the conditional probability table is determined based on the prior probability of the nodes to convert the simulated accident scenarios into constraint conditions;

[0107] The probability of a node propagating to an adjacent node is determined based on the maximum a posteriori probability estimation method under constraints. The expression is:

[0108]

[0109] in, For node N i In new state Under the condition that node N j Transition to a new state The probability of For node N i In new state Under the condition that node N j Transition to a new state The conditional probability of For node N i In new state The prior probability of α is the weight parameter of the prior probability, C is the constraint factor, β is the weight factor of the constraint, and D ij For node N i With node N j The distance, S j For node N j All possible state values ​​of For node N i In new state Under the condition that node N j The probability of being in various possible states;

[0110] For each simulated accident scenario, the paths with the highest probability of propagating to adjacent nodes are selected from the nodes and combined to obtain the risk propagation path in the urban network topology.

[0111] In this embodiment, the method of constructing a risk assessment model to assess the spatiotemporal impact range of each node and performing weighted summation on the assessment results of each node to obtain the city safety risk assessment result includes:

[0112] A risk assessment model was constructed based on a three-dimensional convolutional neural network. The historical recovery time, loss degree, and the spatial and temporal impact range of historical accidents were used as training sets to train the model. The predicted recovery time and loss degree were used as evaluation indicators.

[0113] Calculate the importance of each node, the expression is:

[0114]

[0115] Among them, I i is the importance of node i, K i is the degree of node i in the topological network structure, q i is the resource flow of node i, C i is the expected recovery time of node i, η is the risk resistance sensitivity coefficient, R i is the risk resistance assessment value of node i, μ R and σ R are the mean and standard deviation of the risk resistance assessment values ​​of all nodes, θ is the node association strength weight coefficient, S i is the mean of the association degree between node i and the rest of the nodes;

[0116] The evaluation index of each node is weighted and summed based on the importance of the node to obtain the risk assessment value of the entire urban topological network structure.

[0117] Figure 2 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 2 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0118] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0119] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0120] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a digital device for urban safety risk assessment standards at the logical level. The processor executes the program stored in the memory and is specifically used to execute any of the aforementioned digital methods for urban safety risk assessment standards.

[0121] The present invention can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, including a central processing unit, a network processor (Network Processor, NP), etc.; it can also be a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor to be executed, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0122] An embodiment of the present application also proposes a computer-readable storage medium, which stores one or more programs, and the one or more programs include instructions. When the instructions are executed by an electronic device including multiple applications, they execute any of the aforementioned digital methods of urban safety risk assessment standards.

[0123] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A digital method for urban safety risk assessment standards, characterized in that: The following steps are involved: Identify risk factors from historical accident time series data and obtain a risk factor set; The historical accident time series data includes pre-disaster data, disaster data and post-disaster data; Performing correlation analysis on the risk factor set to obtain a risk factor correlation function; Constructing a city network topology structure based on the risk factor correlation function, and determining a priori probability of each node according to the similarity between each node and the pre-disaster data; Generate a simulated accident scenario, add a disturbance item to simulate a sudden risk event, and obtain a propagation path of the risk in the urban network topology structure in combination with the prior probability of each node; The spatiotemporal influence range of each node traversed by the propagation path is obtained, a risk assessment model is constructed to evaluate the spatiotemporal influence range of each node, and the evaluation results of each node are weighted and summed to obtain the urban safety risk assessment result.

2. According to claim 1, a digital method for urban safety risk assessment standards is characterized in that: The method of identifying risk factors for historical accident time series data and obtaining a risk factor set includes: Extract environmental features from pre-disaster data, extract disaster characteristics from disaster data, and extract loss degree characteristics from post-disaster data, and combine them as a set of urban safety risk features; Calculate the degree of correlation between features, the expression is: Among them, X i and Y j are the i-th item of feature sequence X and the j-th item of feature sequence Y, respectively. m and n are the number of feature sequences X and Y, respectively. α ij and β ij X i and Y j The linear correlation weight and nonlinear correlation weight between ij For X i The positive covariance power adjustment coefficient, μ ij Y j The positive covariance power adjustment coefficient, δ ij Y j The absolute value of the covariance power adjustment coefficient, Cov(X i ,Y j ) is X i and Y j The covariance between and X i and Y j The norm of ; If the correlation between environmental characteristics and disaster characteristics, and between environmental characteristics and loss degree characteristics are greater than the preset threshold, the corresponding environmental characteristics will be used as risk factors to obtain a risk factor set.

3. According to claim 1, a digital method for urban safety risk assessment standards is characterized in that: The method of performing correlation analysis on the risk factor set to obtain a risk factor correlation function includes: A support vector machine model was constructed, and the risk factors for correlation analysis were used as input and output to train the model. The kernel function of the model was used to represent the correlation function, and the expression was: Where N is the number of support vectors, α i is the positive weight of the i-th support vector on the prediction result, is the negative weight of the i-th support vector on the prediction result, γ is the width of the kernel function, x is the current risk factor input vector, x i is the risk factor input vector of the i-th support vector, and b is the bias term.

4. According to claim 1, a digital method for urban safety risk assessment standards is characterized in that: The method for constructing a city network topology structure based on the risk factor association function comprises: Each functional area in the city is defined as a node. The node types include geographical area nodes, infrastructure nodes, and personnel gathering nodes. Each node is assigned an attribute value based on the risk element data. The function of determining the edge between nodes based on the multilayer perceptron neural network is as follows: the number of neurons in the input layer and the output layer are respectively set to the number of risk factors of the two nodes, and the correlation function between the risk factors is used as the mapping function of the corresponding node in the hidden layer of the neural network to obtain the edge function; Establish a data collection system to continuously collect risk element data, update node attribute values ​​and edge functions based on the determined risk element data and risk element association functions, and dynamically adjust the network topology.

5. According to claim 1, a digital method for urban safety risk assessment standards is characterized in that: The method for determining the prior probability of each node according to the similarity between each node and the pre-disaster data includes: Get the pre-disaster data corresponding to the node and calculate the similarity between the node and the pre-disaster data. The expression is: Among them, s i is the similarity between the i-th node and the pre-disaster data, k is the number of features, ω j is the weight of the jth feature, x j and ij are the jth eigenvalue of the node and the jth eigenvalue of the i-th sample in the historical pre-disaster data, max(ξ j ) and min(ξ j ) are the maximum and minimum values ​​of the jth feature in the historical pre-disaster data samples, ∈ is a smoothing term, which is 0.01; Calculate the prior probability of each node, the expression is: Among them, P(s i ) is the prior probability of the ith node, e is a constant, μ is the mean of the similarity distribution, σ is the standard deviation of the similarity, α1 is the amplitude adjustment parameter of the sine function, β1 is the period adjustment parameter of the sine function, γ1 is the translation parameter of the sine function, n is the total number of nodes, s j is the similarity between the jth node and the pre-disaster data.

6. According to claim 1, a digital method for urban safety risk assessment standards is characterized in that: The method of generating a simulated accident scenario and adding a disturbance item to simulate an unexpected risk event includes: A deep convolutional generative adversarial network is selected to construct a generator network and a discriminator network, which are trained based on historical accident case data. The latent space distribution of historical accident data is input into the generator network to obtain a set of simulated accident scenarios. Add disturbance terms to simulate accident scenarios. The disturbance term expression is: Among them, t is the time variable, m is the total number of disturbance terms, P i is the weight of the ith disturbance term, χ i is the fluctuation adjustment coefficient of the ith disturbance term, ψ i is the fluctuation frequency parameter of the ith disturbance term, t 0i is the starting time of the ith disturbance term, ω i is the Gaussian attenuation coefficient of the ith disturbance term, H(·) is a step function, when t≥t 0i When H(tt 0i )=1, indicating that the disturbance term begins to take effect, otherwise the disturbance term has no effect.

7. The digital method for urban safety risk assessment standard according to claim 1 is characterized in that: The method of obtaining the propagation path of risk in the urban network topology structure in combination with the prior probability of each node includes: The nodes in the urban topology network are used as nodes of the Bayesian network, and the conditional probability table is determined based on the prior probability of the nodes to convert the simulated accident scenarios into constraint conditions; The probability of a node propagating to an adjacent node is determined based on the maximum a posteriori probability estimation method under constraints. The expression is: in, At node N i In new state Under the condition that node N j Transition to a new state The probability of For node N i In new state Under the condition that node N j Transition to a new state The conditional probability of For node N i In new state The prior probability of α is the weight parameter of the prior probability, C is the constraint factor, β is the weight factor of the constraint, and D ij For node N i With node N j The distance, S j For node N j All possible state values ​​of For node N i In new state Under the condition that node N j The probability of being in various possible states; For each simulated accident scenario, the paths with the highest probability of propagating to adjacent nodes are selected from the nodes and combined to obtain the risk propagation path in the urban network topology.

8. The digital method for urban safety risk assessment standard according to claim 1 is characterized in that: The method of constructing a risk assessment model to assess the spatiotemporal impact range of each node and performing weighted summation on the assessment results of each node to obtain the urban safety risk assessment result includes: A risk assessment model was constructed based on a three-dimensional convolutional neural network. The historical recovery time, loss degree, and the spatial and temporal impact range of historical accidents were used as training sets to train the model. The predicted recovery time and loss degree were used as evaluation indicators. Calculate the importance of each node, the expression is: Among them, I i is the importance of node i, K i is the degree of node i in the topological network structure, q i is the resource flow of node i, C i is the expected recovery time of node i, η is the risk resistance sensitivity coefficient, R i is the risk resistance assessment value of node i, μ R and σ R are the mean and standard deviation of the risk resistance assessment values ​​of all nodes, θ is the node association strength weight coefficient, S i is the mean of the association degree between node i and the rest of the nodes; The evaluation index of each node is weighted and summed based on the importance of the node to obtain the risk assessment value of the entire urban topological network structure.

9. An electronic device, comprising: processor; as well as A memory arranged to store computer executable instructions, which when executed cause the processor to perform the method according to any one of claims 1 to 8.

10. A computer-readable storage medium storing one or more programs, which, when executed by an electronic device including a plurality of application programs, enables the electronic device to execute the method according to any one of claims 1 to 8.

Images