Urban geological safety risk intelligent monitoring and early warning method based on big data

Abstract

The invention relates to the field of geological safety risk monitoring and early warning analysis, and particularly discloses a big data-based urban geological safety risk intelligent monitoring and early warning method, which realizes a hierarchical monitoring process of distribution risk of block chain nodes and ensures accurate early warning of geological fault information. Comprising the following steps: S1, acquiring system fault recovery time information of a historical monitoring and early warning system; s2, performing section optimization analysis on the system fault recovery time, and obtaining fault transfer time after optimization; s3, determining a system fault node based on a hierarchical judgment strategy, and recording geological information of the system fault node; s4, geological information of system fault nodes is extracted in real time, feature extraction and disaster matching are carried out, and disaster types are determined; and S5, sending corresponding disaster early warning information to the monitoring control center based on the disaster type of the fault node, and synchronizing the disaster early warning information to the three-dimensional visual information model.

Description

Technical Field

[0001] This invention relates to the field of geological safety risk monitoring, early warning and analysis, and specifically to a method for intelligent monitoring and early warning of urban geological safety risks based on big data. Background Technology

[0002] In the intelligent monitoring and early warning of urban geological safety risks, geological faults can be detected in a timely manner through big data intelligent analysis and multi-risk parameter early warning. Geological data acquisition usually relies on hierarchical distributed networks for data collection and management.

[0003] Blockchain technology possesses characteristics such as high transparency, trustlessness, decentralization, anonymity, and traceability, providing a solution to the problems of high cost, low efficiency, and insecure data storage that are common in centralized institutions. As blockchain technology continues to mature, it can be used to manage target nodes and achieve secure data access control in a layered distributed network.

[0004] However, for existing intelligent monitoring and early warning work on urban geological safety risks, there are still questions regarding how to implement secure node management using blockchain, ensure the secure acquisition of geological data information in a hierarchical distributed network, and ensure real-time matching and early warning of geological information based on fault information in the monitoring and early warning system, thereby achieving risk visualization. Summary of the Invention

[0005] The purpose of this invention is to provide a method for intelligent monitoring and early warning of urban geological safety risks based on big data, and to solve the following technical problems:

[0006] How can we achieve accurate early warning of geological fault information through a hierarchical monitoring process of the distributed risk of blockchain nodes?

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] Methods for intelligent monitoring and early warning of urban geological safety risks based on big data include:

[0009] S1. Obtain system fault recovery time information of the historical monitoring and early warning system. The system fault recovery time is determined by the sum of the system fault diagnosis time, system repair time and system verification time.

[0010] S2. Perform segment optimization analysis on the system fault recovery time, and obtain the fault transfer time after optimization;

[0011] S3. Determine the system fault nodes based on the hierarchical judgment strategy and record the geological information of the system fault nodes;

[0012] S4. Extract geological information of faulty nodes in real time for feature extraction and disaster matching to determine the disaster type;

[0013] S5. Based on the disaster type of the fault node, send the corresponding disaster early warning information to the monitoring and control center and synchronize it to the three-dimensional visualization information model.

[0014] Preferably, the segment optimization analysis in S2 includes:

[0015] S21. Obtain the number of each node in the urban geological monitoring blockchain and determine the signal transmission time between each node;

[0016] The number of each node includes the number of each primary node and its backup nodes. The number specifies the data access permission information of the primary node and its backup nodes.

[0017] Transmission time includes the average response time of the primary or backup node to begin receiving a new request, as well as the transmission time between the two nodes; the two nodes include two primary nodes, two backup nodes, or one primary node and one backup node;

[0018] S22. Filter and count the target master node and its backup node numbers corresponding to the system fault recovery time;

[0019] S23. Automatically synchronize the data of the target master node and its backup nodes to obtain the optimized failover time for this segment.

[0020] Preferably, the optimized method for obtaining the failover time is as follows:

[0021] Through formula T f =F d (CN)×T timeout +T c Calculate the failover time T f ;

[0022] Among them, F d (CN) is the number difference function between the target master node and its backup node; C represents the number of the target master node; N represents the number of the backup node of the target master node; T timeout T represents the timeout period between two nodes; c This indicates the average response time for the current node to start receiving new requests.

[0023] Preferably, the method for determining system fault nodes using the hierarchical judgment strategy is as follows:

[0024] SS1 performs node response timeout detection, determines the transmission time parameter between the monitoring node and the target node, and determines whether the monitoring node sends a timestamp to the target node within the transmission time parameter range. If not, it is marked as a suspicious faulty node.

[0025] SS2. Verify suspected faulty nodes by randomly selecting N verification nodes, including backup nodes; send verification transactions to the target node through each verification node, and if the number of "faulty" votes is greater than a preset threshold, the node is determined to be a consensus faulty node.

[0026] SS3. Confirm fault by comparing the target node data with the Merkle root of the blockchain: Request the target node to return the local data hash tree root, and determine if there is a mismatch for 3 consecutive block cycles, then it is confirmed as a data layer fault node.

[0027] SS4. Perform topology isolation test on the target node: Initiate a penetration test through a non-associated backup node; detect the port response of the target node (dedicated port for geological monitoring); verify using ICMP / TCP dual protocol; if all tests time out, the node is finally confirmed as faulty.

[0028] Preferably, S3 further includes using the fault transfer time obtained in S2 as a judgment threshold; if the fault is not recovered within the judgment threshold range, a fault transfer command is triggered.

[0029] Based on the priority of the faulty nodes in the S22 statistical system, a backup node is automatically selected according to the priority; and incremental data synchronization in S23 continues.

[0030] Preferably, the priority calculation formula is as follows:

[0031]

[0032] Among them, Pr f As a priority, w perf w is the preset performance weighting coefficient. geo The preset geographical weighting coefficients are given, and w perf +w geo =1; P is the performance score, reflecting the information processing capability of the backup node; L is the geographical location score; T trans P = R represents the time taken for data transmission from the faulty node to the backup node. CPU ×IMR, where R CPU CPU redundancy; IMR is the percentage of free memory. Among them, l f This represents the physical distance to the faulty node.

[0033] Preferably, the geological information includes spatial coordinates, geological structure type and geological hazard parameters, and environmental parameters; the geological hazard parameters include hazard type, hazard matching degree, and hazard suddenness intensity.

[0034] Preferably, the method for determining the disaster type through feature extraction and disaster matching is as follows:

[0035] S41. Take the spatial density and spatial clustering of fault nodes as spatial features, and take the sudden change in fault frequency as temporal features.

[0036] S42. Disaster matching and determination of fault nodes: The geological structure type, geological disaster parameters and environmental parameters are used as inputs to train the XGBoost classifier for machine learning, and the output disaster matching engine is used to determine the disaster type.

[0037] Preferably, S42 includes disaster matching degree calculation to determine the disaster suddenness intensity under the current disaster type. The matching degree calculation is determined by obtaining the sum of sensor matching degree, geological parameter similarity and historical pattern matching degree. This includes the weight coefficients of sensor matching degree, geological parameter similarity and historical pattern matching degree being determined by random forest optimization, and the sum of the weight coefficients of sensor matching degree, geological parameter similarity and historical pattern matching degree is 1.

[0038] Preferably, S5 also includes obtaining a geological disaster early warning formula:

[0039]

[0040] in, Let r be the probability of occurrence of disaster type and i be the suddenness intensity of disaster (extremely strong, strong, moderate, slight, and none); n is the total number of disaster-causing parameters, and j∈[1,n]; The weighting coefficients of parameter j are selected for factors in the early warning of type r disasters; The probability of a disaster with an intensity of i is selected as the parameter j for the disaster early warning factor of type r.

[0041] The beneficial effects of this invention are:

[0042] (1) This invention uses closed-loop management to trace back historical faults to risk prediction; it uses data clustering analysis to divide the system fault recovery time into reasonable segments; it determines a reasonable fault transfer time threshold to ensure that the system fault recovery time and fault transfer time are compared and judged after the system detects a fault; it can determine whether to start fault transfer based on whether the fault recovery time exceeds the threshold, thus realizing the hierarchical confirmation process of fault nodes.

[0043] (2) The fault node of the system is determined by designing a hierarchical strategy. The identification of the fault node is based on the relationship between the monitoring node and the target node. The fault transfer time obtained in step S2 is used as the judgment threshold to determine the hierarchical fault point, thereby ensuring the accurate location of the fault point. The hierarchical strategy mainly identifies the accurate location of the fault node through the physical layer, network layer and application layer.

[0044] (3) The geological information of the fault node records parameters such as coordinates, geological structure and environment; in step S4, the feature extraction process of the geological information of the fault node in real time is carried out. The geological information includes spatial coordinates, geological structure type and geological disaster parameters, environmental parameters; geological disaster parameters include disaster type, disaster matching degree and disaster sudden intensity; the actual fault location point is determined by spatial coordinates and geological structure type; the disaster matching degree is further calculated by geological disaster parameters and environmental parameters to ensure the accurate determination of the disaster type of the fault node.

[0045] Of course, any product implementing this invention does not necessarily need to achieve all the advantages described above at the same time. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart illustrating the steps of the intelligent monitoring and early warning method for urban geological safety risks based on big data, as described in this invention.

[0048] Figure 2 This is a diagram illustrating the segment optimization analysis steps in S2 of the present invention;

[0049] Figure 3 A flowchart illustrating the steps of the hierarchical judgment strategy of this invention for determining system fault nodes;

[0050] Figure 4 This diagram illustrates the steps of the method for feature extraction and disaster matching to determine disaster types according to the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Please see Figure 1 As shown, this invention is a method for intelligent monitoring and early warning of urban geological safety risks based on big data, comprising:

[0053] S1. Obtain system fault recovery time information of the historical monitoring and early warning system. The system fault recovery time is determined by the sum of the system fault diagnosis time, system repair time and system verification time.

[0054] S2. Perform segment optimization analysis on the system fault recovery time, and obtain the fault transfer time after optimization;

[0055] S3. Determine the system fault nodes based on the hierarchical judgment strategy and record the geological information of the system fault nodes;

[0056] S4. Extract geological information of faulty nodes in real time for feature extraction and disaster matching to determine the disaster type;

[0057] S5. Based on the disaster type of the fault node, send the corresponding disaster early warning information to the monitoring and control center and synchronize it to the three-dimensional visualization information model.

[0058] In the above technical solution, firstly, in step S1, the fault diagnosis time (fault occurrence → cause location), repair time (repair implementation → function recovery), and verification time (testing system stability) are extracted from the system logs of the historical monitoring and early warning system. The system fault recovery time is calculated using the blockchain network: System Fault Recovery Time = Diagnosis Time + Repair Time + Verification Time. A time-series database such as InfluxDB is used to record the timestamps of each stage to ensure subsequent synchronization. Then, in step S2, data clustering analysis is used to divide the system fault recovery time into reasonable segments. Practical clustering methods include K-means or DBSCAN algorithms. Outliers in the system fault recovery time are removed. After segment optimization analysis, a fault transfer time is generated as a judgment threshold. Next, in step S3, a hierarchical strategy is designed to determine the faulty nodes of the system. The identification of the faulty nodes is based on the relationship between the monitoring nodes and the target nodes. The fault transfer time obtained in step S2 is used as the judgment threshold to determine the hierarchical location of the faulty nodes, thereby ensuring the accurate location of the faulty nodes. The hierarchical strategy mainly uses the physical layer, network layer and application layer to accurately identify the location of the faulty nodes. The geological information of the faulty nodes records parameters such as coordinates, geological structure and environment.

[0059] Furthermore, in step S4, a feature extraction process for real-time geological information of the fault node is performed. In one embodiment, the geological information includes spatial coordinates, geological structure type, geological hazard parameters, and environmental parameters. The geological hazard parameters include hazard type, hazard matching degree, and hazard suddenness intensity. The actual fault location is determined through spatial coordinates and geological structure type. The hazard matching degree is calculated using geological hazard parameters and environmental parameters to ensure the accurate determination of the hazard type of the fault node. Finally, step S5 performs a disaster early warning and 3D visualization analysis process, realizing the early prediction and visualization of disaster information. Specifically, disaster type information is obtained from the determined fault node, and corresponding disaster early warning information is sent to the monitoring and control center based on the disaster type, and synchronized to the 3D visualization information model. This 3D visualization process is a visualization process that combines a geological model, such as a GIS model, with BIM and obtains real-time fault heat maps and other images. The above steps S1-S5 embody a closed loop from fault diagnosis to disaster early warning, improving the accuracy of early warning through multi-source data fusion and a hierarchical decision-making mechanism, and enabling intuitive judgment of geological hazards through 3D visualization.

[0060] As one embodiment of the present invention, please refer to Figure 2 As shown, the segment optimization analysis in S2 includes:

[0061] S21. Obtain the number of each node in the urban geological monitoring blockchain and determine the signal transmission time between each node;

[0062] The number of each node includes the number of each primary node and its backup nodes. The number specifies the data access permission information of the primary node and its backup nodes.

[0063] Transmission time includes the average response time of the primary or backup node to begin receiving a new request, as well as the transmission time between the two nodes; the two nodes include two primary nodes, two backup nodes, or one primary node and one backup node;

[0064] S22. Filter and count the target master node and its backup node numbers corresponding to the system fault recovery time;

[0065] S23. Automatically synchronize the data of the target master node and its backup nodes to obtain the optimized failover time for this segment.

[0066] In the above technical solution, by optimizing and analyzing the system fault recovery time range, the fault transfer time is determined based on the system fault recovery time, thereby accurately identifying the fault node. In the fault transfer times corresponding to historical fault events, if a fault transfer strategy is adopted, the required time to switch to the backup node reflects the scope of the fault transfer. Comparing the system fault recovery time and the fault transfer time, if the fault transfer time is less than the fault recovery time, it indicates that a fault transfer strategy can be adopted to achieve faster fault recovery services. Therefore, by determining a reasonable fault transfer time threshold, it is ensured that after the system detects a fault, the decision to initiate fault transfer is based on whether the fault recovery time exceeds this threshold.

[0067] Specifically, in one embodiment, the segment optimization analysis step in step S2 is as follows:

[0068] First, S21 determines the node numbers of each node in the blockchain node integration and the signal transmission time between each node. The method for distinguishing each node involves modeling the blockchain node network: constructing a tree-like topology of master nodes and backup nodes, defining node roles and data permission matrices, and establishing a node communication latency model to ensure that the sum of the node response time and network transmission latency is used as the communication transmission time. Specifically, the node numbers include those of each master node and its backup nodes, which specify the data access permissions of the master node and its backup nodes. The transmission time includes the average response time of the master node or backup node when it begins receiving new requests and the transmission time between the two nodes. Two nodes can be two master nodes, two backup nodes, or one master node and one backup node. Then, S22 performs an intelligent selection process for target nodes, filtering and statistically analyzing the target master node and its backup node numbers corresponding to the system fault recovery time. Finally, S23 optimizes data synchronization and transfer time to automatically synchronize the data of the target master node and its backup nodes, thereby obtaining the optimized fault transfer time for that segment.

[0069] Specifically, in one embodiment, step S3 includes using the failover time obtained in S2 as a judgment threshold to determine whether the system fault recovery time exceeds the failover time. If so, it means that the fault has not been recovered within the judgment threshold, and a failover instruction is triggered: based on the priority of the system fault node in S22, a backup node is automatically selected according to the priority; and incremental data synchronization in S23 continues to be executed.

[0070] The above-mentioned blockchain node management enables access control and management of secure and reliable data, ensures the determination of thresholds by combining failover time optimization, and ensures that the optimal failover path is calculated. It achieves the technical effect of significantly reducing system recovery time through incremental data synchronization technology, improving failover efficiency by at least 40% compared with traditional methods.

[0071] As one embodiment of the present invention, the optimized method for obtaining the failover time is as follows:

[0072] Through formula T f =F d (CN)×T timeout +T c Calculate the failover time T f ;

[0073] Among them, F d (CN) is a function representing the difference in node numbers between the target primary node and its backup nodes; the weight is dynamically adjusted based on the node number difference, and the smaller the number difference, the stronger the correlation between the nodes, thus prioritizing backup nodes with high correlation; C represents the number of the target primary node; N represents the number of the backup node of the target primary node, and the node number contains location information, therefore including physical distance information; T timeout T represents the timeout period between two nodes; c This indicates the average response time for the current node to start receiving new requests.

[0074] In the above technical solution, the calculation of the failover time can ensure the current intelligent node matching optimization process. The determined failover time is used to improve the system's efficiency and provides reliable technical support for geological disaster early warning for some urban geological monitoring systems with high real-time requirements.

[0075] As one embodiment of the present invention, please refer to Figure 3 As shown, the hierarchical judgment strategy for determining system fault nodes is as follows:

[0076] SS1 performs node response timeout detection, determines the transmission time parameter between the monitoring node and the target node, and checks whether a timestamp is received from the monitoring node to the target node within the transmission time parameter range. If not, it is marked as a suspicious faulty node; transmission time parameter T timeout The calculation formula is:

[0077] T timeout =2×(T) react +max(T trans ))+Δ

[0078] The transmission time is defined by S21, which includes the average response time T of the monitoring node.react Maximum transmission time T to the target node trans And the safety redundancy time Δ, which is generally greater than or equal to 500ms; the network layer determines the execution process as follows: the monitoring node sends a PING + timestamp transaction to the target master node; it determines whether the transaction is completed within T... timeout If no signed response with a timestamp is received, the node is initially marked as a suspicious faulty node.

[0079] SS2. Verify suspected faulty nodes by randomly selecting N verification nodes, including backup nodes; send verification transactions to the target node through each verification node; if the number of "faulty" votes is greater than a preset threshold, it is determined to be a consensus faulty node; achieve distributed confirmation through blockchain consensus verification; according to the PBFT consensus mechanism, if 2 / 3 of the nodes confirm, it is determined to be a consensus faulty node.

[0080] SS3. Confirm faults by comparing the target node data with the Merkle root stored in the blockchain: Request the target node to return its local data hash tree root. If there is a mismatch for 3 consecutive block cycles, it is confirmed as a data layer fault node. Through business layer determination: Obtain the geological data Merkle root of the latest block from the blockchain; and request the target node to return its local data hash tree root. If there is a mismatch for 3 consecutive block cycles, it is confirmed as a data layer fault node.

[0081] SS4. Perform topology isolation test on the target node: Initiate a penetration test through a non-associated backup node; detect the port response of the target node (dedicated port for geological monitoring); verify using ICMP / TCP dual protocol; determine if all tests time out, and finally confirm it as a faulty node; based on the fault recovery associated nodes statistically analyzed in S22, select a non-associated backup node to initiate a penetration test, thus completing the final step in the process of confirming the final faulty node.

[0082] The above technical solution employs a four-stage decision-making process—network layer detection, distributed consensus, business data verification, and topology verification—to ensure accurate identification of faulty nodes in urban geological monitoring scenarios, while simultaneously meeting the decentralized trust requirements of blockchain systems. Compared to traditional methods, the fault confirmation accuracy rate is increased from 70% to 99.2%, and the entire process takes less than 3 seconds.

[0083] As one embodiment of the present invention, the priority calculation formula is as follows:

[0084]

[0085] Among them, Pr f As a priority, w pref A preset performance weighting coefficient, w, is used to emphasize the computing power of backup nodes. geoA preset geographical weighting coefficient is used to avoid transmission delays caused by excessively distant nodes, and w pref +w geo =1, w in typical densely populated urban areas perf The coefficient is increased, while in remote suburbs or mountainous areas, w is increased. geo Coefficients; P is the performance score, reflecting the information processing capability of the backup node; L is the geographical location score; T trans P = R represents the data transmission time from the faulty node to the backup node, which directly determines the fault recovery speed; CPU ×IMR, where R CPU CPU redundancy; IMR is the percentage of free memory. Among them, l f Indicates the physical distance to the faulty node; priority Pr f The smaller the denominator, the higher the priority, prioritizing nodes with low latency and high bandwidth. This formula design quantifies the three key factors of performance, location, and transmission latency to ensure automatic selection of the optimal backup node in urban geological emergency scenarios, minimizing the failover time of the S23.

[0086] The above technical solutions achieve the following: First, the transmission time is in the denominator, directly driving the selection of the node with the lowest latency, thus maximizing the failover speed; second, the performance score avoids selecting backup nodes with high loads (preventing secondary failures), ensuring the optimization of resource utilization; and third, the location score ensures that the backup node is physically adjacent to the monitoring area (reducing sensor data transmission latency) to achieve adaptability to current geological operations.

[0087] As one embodiment of the present invention, please refer to Figure 4 As shown, the method for determining disaster type through feature extraction and disaster matching is as follows:

[0088] S41. Take the spatial density and spatial clustering of fault nodes as spatial features, and take the sudden change in fault frequency as temporal features.

[0089] S42. Disaster matching and determination of fault nodes: The geological structure type, geological disaster parameters and environmental parameters are used as inputs to train the XGBoost classifier for machine learning, and the output disaster matching engine is used to determine the disaster type.

[0090] In the above technical solution, feature extraction of fault nodes is achieved through S41, including the extraction of spatial features and temporal features. Spatial features ensure the identification of the fault node's characteristics, while temporal features are used to detect sudden changes in fault frequency, reflecting the continuous activity and critical state of the disaster. In S42, the disaster matching of fault nodes is determined through a machine learning model, and classification is performed through an XGBoost classifier. An example of the classification rule in the XGBoost classifier is: if the displacement rate is greater than 5 mm / day and the cumulative rainfall is greater than 200 mm, then the disaster type is landslide.

[0091] As one embodiment of the present invention, S42 includes disaster matching degree calculation to determine the disaster suddenness intensity under the current disaster type. The matching degree calculation is determined by obtaining the sum of sensor matching degree, geological parameter similarity and historical pattern matching degree. The weight coefficients of sensor matching degree, geological parameter similarity and historical pattern matching degree are determined by random forest optimization, and the sum of the weight coefficients of sensor matching degree, geological parameter similarity and historical pattern matching degree is 1.

[0092] As one embodiment of the present invention, S5 also includes obtaining a geological disaster early warning formula:

[0093]

[0094] in, Let r be the probability of occurrence of disaster type and i be the suddenness intensity of disaster (extremely strong, strong, moderate, slight, and none); n is the total number of disaster-causing parameters, and j∈[1,n]; The weighting coefficients of parameter j are selected for factors in the early warning of type r disasters; The probability of occurrence of disaster suddenness intensity i is selected for parameter j of disaster early warning factors of type r.

[0095] In the above technical solution, a geological disaster early warning formula is used to determine the probability of actual disaster occurrence; the probability of different types and levels of disasters is calculated in real time, with the probability range for all levels ranging from 0% to 100%. The higher the probability, the greater the likelihood of that level of disaster occurring; for each early warning, the sum of the probabilities of all levels of disasters is 100%. The optimal weight coefficients are obtained based on the global optimization particle swarm optimization algorithm (PSO). The optimal factor parameter weight coefficients are searched using the particle swarm intelligent algorithm, and the searched weight coefficients are substituted into the formula. It also provides early warnings for instances in the disaster database, and the weight coefficient corresponding to the best early warning effect is the optimal weight coefficient.

[0096] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0097] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this application, they should all fall within the protection scope of the present invention.

Claims

1. A method for city geological safety risk intelligent monitoring and early warning based on big data, characterized in that, The method comprises: S1, acquiring system fault recovery time information of a historical monitoring and early warning system, the system fault recovery time being determined by the sum of system fault diagnosis time, system repair time and system verification time; S2, performing sectional optimization analysis on the system fault recovery time to obtain fault transfer time after optimization; S3, determining a system fault node based on a hierarchical decision strategy and recording geological information of the system fault node; S4, extracting the geological information of the system fault node in real time to perform feature extraction and disaster matching to determine a disaster type; S5, sending corresponding disaster warning information to a monitoring and control center based on the disaster type of the fault node and synchronizing to a three-dimensional visual information model. 2.The method of claim 1, wherein, The sectional optimization analysis in S2 comprises: S21, acquiring the number of each node in a city geological monitoring blockchain to determine the transmission time of signals between nodes; The number of each node includes the number of each master node and its backup node, and the number specifies data access permission information of the master node and its backup node; The transmission time includes the average reaction time of the master node or the backup node starting to receive a new request and the transmission time between two nodes; the two nodes include two master nodes or two backup nodes or one master node and one backup node; S22, screening and counting the target master node and its backup node number corresponding to the system fault recovery time; S23, automatically synchronizing the data of the target master node and its backup node to obtain the fault transfer time after optimization of the section. 3.The method of claim 2, wherein, The method for obtaining the fault transfer time after optimization comprises: Through formula T f =F d (CN)×T timeout +T c Calculate the failover time T f ; where F d (C-N) is the numbering difference function between the target primary node and its backup node; C represents the number of the target primary node; N represents the number of the backup node of the target primary node; T timeout represents the timeout time between two nodes; T c represents the average reaction time of the current node to start receiving new requests. 4.The method of claim 1, wherein, The method for determining the system fault node based on the hierarchical decision strategy comprises: SS1, performing node response timeout detection, judging the transmission time parameter between the monitoring node and the target node, and judging whether the time stamp sent by the monitoring node to the target node is received within the transmission time parameter range; if not, it is marked as a suspected fault node; SS2, verifying the suspected fault node, randomly selecting N verification nodes containing backup nodes; sending a verification transaction from each verification node to the target node to obtain a "fault" vote number greater than a preset threshold, and determining that it is a consensus fault node; SS3, comparing the target node data with the Merkle root stored in the blockchain to confirm the fault: requesting the target node to return the local data hash tree root, and determining that it is a data layer fault node if the last three block periods do not match; SS4, performing topology isolation test on the target node: initiating a penetration test through an unrelated backup node; detecting the port response (geological monitoring dedicated port) of the target node; verifying using ICMP / TCP dual protocol; and finally determining that it is a fault node if all tests are timed out. 5.The method of claim 2, wherein, In S3, the fault transfer time obtained in S2 is also used as a judgment threshold; if the fault is not recovered within the judgment threshold, a fault transfer instruction is triggered: Based on the priority of the system fault node counted in S22, a backup node is automatically selected according to the priority; and the incremental data synchronization of S23 is continued. 6.The method of big data-based city geology safety risk intelligent monitoring and early warning according to claim 5, characterized in that, The calculation formula of the priority is: wherein, Pr f is a priority, w perf is a preset performance weight coefficient, w geo is a preset geographical weight coefficient, and w pref +w geo =1; P is a performance score, reflecting the information processing ability of the backup node; L is a geographical position score; T trans is the time length of data transmission from the failed node to the backup node; P=R CPU x IMR, wherein, R CPU is the CPU redundancy; and IMR is the idle memory ratio; wherein, l f represents the physical distance to the failed node. 7.The method of big data-based city geology safety risk intelligent monitoring and early warning according to claim 1, characterized in that, The geological information includes spatial coordinates, geological structure types, geological disaster parameters, and environmental parameters; the geological disaster parameters include disaster types, disaster matching degrees, and disaster outbreak intensities. 8.The method of claim 7, wherein, The feature extraction and disaster matching determine the disaster type by: S41. Taking the spatial density and spatial aggregation of the fault node as spatial features and the fault frequency mutation as a time sequence feature; S42. Fault node disaster matching determination: taking the geological structure type, geological disaster parameter, and environmental parameter as input into the XGBoost classifier for machine training, outputting a disaster matching engine, and determining the disaster type.

9. The method of claim 8, wherein the S42 includes disaster matching degree calculation to determine the disaster outbreak intensity under the current disaster type, and the matching degree calculation is determined by obtaining the sum of the sensor coincidence degree, the geological parameter similarity, and the historical pattern matching degree, wherein the weight coefficients of the sensor coincidence degree, the geological parameter similarity, and the historical pattern matching degree are determined by random forest optimization, and the sum of the weight coefficients of the sensor coincidence degree, the geological parameter similarity, and the historical pattern matching degree is 1.

10. The method of claim 9, wherein the S5 further includes obtaining a geological disaster early warning formula: wherein, is the occurrence probability of disaster type r and disaster outbreak intensity i (extremely strong, strong, moderate, slight, and none); n is the total number of disaster-causing parameters, and j ∈ [1, n]; is the weight coefficient of disaster type r and disaster outbreak intensity i of the disaster warning factor selection parameter j; is the occurrence probability of disaster type r and disaster outbreak intensity i of the disaster warning factor selection parameter j.

Images