Waterlogging early warning information processing method and system and electronic equipment
Through the combination of text embedding model and large language model, basic monitoring and simulation forecast data are processed in real time, and accurate flood warning information is generated, which solves the problem of untimely update of flood warning information in the existing technology, and realizes automatic warning release and loss reduction .
Patent Information
- Application Number
- CN202411090349.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-09
- Publication Date
- 2024-11-26
AI Technical Summary
The existing technology cannot update the flood warning in real time in the context of urbanization and climate change, resulting in reduced accuracy and deviation of early warning information and unable to adapt to urban changes.
By obtaining current basic monitoring data and simulation forecast data, using text embedding models and large language models for cluster analysis and early warning message optimization, to generate real-time and accurate flooding warning information.
Automatic warning release has been achieved, knowledge acquisition and utilization efficiency has been improved, disaster losses have been reduced, and existing solutions have been optimized through historical data to adapt to climate change and land use cover changes.
Smart Images

Figure CN119025949A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular to a method and system for processing waterlogging warning information, and electronic equipment. Background Art
[0002] In recent years, urbanization has developed rapidly. Facing the impact of climate change and land use change caused by high-intensity development, the development of urban hydrology and water system science faces new opportunities and challenges. It mainly manifests itself in changes in the hydrological cycle. Urban construction has squeezed out the original ecosystem, leading to wetland degradation and lake shrinkage, which further causes urban waterlogging. In some technical solutions, numerical results can be detected or predicted by equipment monitoring or numerical simulation, and waterlogging warning release messages can be generated by comparing the known rule base; however, this technical process has the disadvantage that the rule base is outdated and the waterlogging warning release messages cannot be updated in real time. The warning messages issued by traditional flood control and waterlogging emergency plans cannot be corrected in real time with various urban changes, resulting in a decrease in the accuracy of waterlogging warning release messages issued by emergency plans after a period of time, and deviations in warning information. Summary of the invention
[0003] In view of this, the present invention proposes a waterlogging warning information processing method and system, and electronic equipment, which can provide accurate waterlogging warning information in real time.
[0004] The technical solution of the present invention is implemented as follows: The present invention provides a method for processing waterlogging warning information, the method comprising:
[0005] Acquire current basic monitoring data and current simulation forecast data; wherein the current basic monitoring data is used to describe the real-time water condition data currently collected; and the current simulation forecast data is used to describe the simulated water condition data corresponding to a preset time period in the future after the current basic monitoring data is acquired;
[0006] Using a text embedding model, clustering the current basic monitoring data, the current simulation forecast data, and the set of historical monitoring data and historical simulation forecast data is performed to determine the target clustering categories corresponding to the current basic monitoring data and the current simulation forecast data, and the initial warning release message corresponding to the target clustering category; wherein the initial warning release message is determined by the same target clustering category corresponding to the historical monitoring data and the set of warning release messages;
[0007] The historical warning release message and the initial warning release message are input into the trained large language model, and the target warning release message output by the large language model for assisting decision making is obtained.
[0008] On the basis of the above technical solution, preferably, before clustering the current basic monitoring data, the current simulation forecast data, and the set of historical monitoring data and historical simulation forecast data, the method further includes:
[0009] According to the historical basic monitoring data and historical simulation forecast data corresponding to each historical time point, a collection of historical monitoring data and historical simulation forecast data is constructed;
[0010] Obtaining historical warning release messages corresponding to the historical basic monitoring data and historical simulation forecast data at each historical time point, optimizing the historical warning release messages to obtain corresponding optimized warning release messages, and constructing an optimized warning message set;
[0011] The historical monitoring data, the historical simulation forecast data set, and the optimized early warning release message set are combined and processed based on historical time points to establish the historical monitoring data and early warning release message set.
[0012] On the basis of the above technical solution, the optimization processing of the historical warning release message to obtain the corresponding optimized warning release message includes:
[0013] Establish manual analysis tasks for historical warning release messages at each historical time point, perform data analysis on the historical basic monitoring data and waterlogging data corresponding to each historical time point, optimize the historical warning release messages based on the data analysis results, and determine the corresponding optimized warning release messages.
[0014] On the basis of the above technical solution, preferably, the optimization warning release message set includes optimization warning release messages at multiple different historical time points, expressed as:
[0015] Y={(y 11 ,y 12 ,…),(y 21 ,y 22 …),…,(y m1 ,y m2 , …)}
[0016] Among them, Y is the set of optimized warning release messages; m1 Represents the first message at the mth moment, and the number of subsets at each moment is different.
[0017] On the basis of the above technical solution, preferably, the optimized warning release message set includes: a historical monitoring data and historical simulation forecast data set X, a historical monitoring data and warning release message set M;
[0018] Where X = {(x 11 , x 12 , …, x1n ), (x 21 ,x 22 …, x 2n ),…,(x m1 , x m2 ,…,x mn )}; where () represents the subset at the mth moment; x mn Represents the nth type of data at the mth moment;
[0019] M={(x1:y1),(x2:y2)…,(x m :y m )}; x m Represents the historical monitoring data subset at the mth moment (x m1 , x m2 , …, x mn );y m Represents the optimized warning release message subset at the mth moment (y m1 ,y m2 , …).
[0020] On the basis of the above technical solution, preferably, the text embedding model is used to cluster the basic monitoring data, the simulated forecast data, and the historical monitoring data and the historical simulated forecast data set, and the target clustering category corresponding to the basic monitoring data and the simulated forecast data, and the initial warning release message corresponding to the target clustering category is determined, including:
[0021] Combine the current basic monitoring data with the current simulation forecast data The historical monitoring data and historical simulation forecast data set X are simultaneously input into the text embedding model for clustering analysis to determine the clustering categories of each subset in the set, expressed as Among them, C is the cluster category set; Represents the current basic monitoring data and the current simulation forecast data set Cluster category; c m Represents the clustering categories of each subset in the historical monitoring data and historical simulation forecast data set X;
[0022] Based on the current basic monitoring data and current simulation forecast data set The mapping relationship between the historical monitoring data and the historical simulation forecast data set X is used to determine the current basic monitoring data and the current simulation forecast data set Clustering categories
[0023] According to clustering categories and historical monitoring data subsets (c1, c2, ..., c m )’s intersection determines the same cluster category c′;
[0024] According to the mapping relationship between the historical monitoring data and the historical simulation forecast data set X and the cluster category set C, the historical warning data subset x′ with the same cluster category c′ is determined;
[0025] Based on the correspondence between the historical monitoring data set and the optimized warning release message set, the search result y′ in the optimized warning release message set with the same clustering category is determined in the historical monitoring simulation data and the warning release message set M as the initial warning release message.
[0026] On the basis of the above technical solution, preferably, the current basic monitoring data and the current simulation forecast data are combined The historical monitoring data and the historical simulation forecast data set X are simultaneously input into the text embedding model for cluster analysis. The text embedding model is input for cluster analysis, including:
[0027] Randomly select k clusters as the initial cluster center P0;
[0028] For the current basic monitoring data and the current simulation forecast data set For each data point in the historical monitoring data and historical simulation forecast data set X, calculate its distance from each cluster center and assign it to the cluster center category C with the closest distance k , is the set of data points of the kth cluster;
[0029] For k clusters, recalculate each cluster center; wherein the updated cluster center is determined based on the mean of all data points in the cluster;
[0030] Iterate based on the updated cluster centers until the preset termination condition is met to complete clustering.
[0031] On the basis of the above technical solution, preferably, the current basic monitoring data includes: at least one of rainfall, water level and water depth data;
[0032] The current simulation forecast data includes: at least one of the rainfall, water level and water depth data estimated based on the basic monitoring data.
[0033] More preferably, the present invention provides a waterlogging warning information processing system, the system comprising:
[0034] A data acquisition module, used to acquire current basic monitoring data and current simulation forecast data; wherein the current basic monitoring data is used to describe the real-time water condition data currently collected; and the current simulation forecast data is used to describe the simulated water condition data corresponding to a preset time period in the future after the current basic monitoring data is acquired;
[0035] A text embedding model processing module, used to cluster the current basic monitoring data, the current simulation forecast data, and the historical monitoring data and historical simulation forecast data sets using a text embedding model, and determine the target clustering categories corresponding to the current basic monitoring data and the current simulation forecast data, and the initial warning release message corresponding to the target clustering category; wherein the initial warning release message is determined by the same target clustering category corresponding to the historical monitoring data and the warning release message set;
[0036] The large language model processing module is used to input the historical warning release message and the initial warning release message into the trained large language model, and obtain the target warning release message output by the large language model for auxiliary decision-making.
[0037] More preferably, the present invention provides an electronic device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the waterlogging warning information processing method described in any one of the above embodiments by executing the executable instructions.
[0038] The waterlogging warning information processing method of the present invention has the following beneficial effects compared with the prior art:
[0039] By collecting the current basic monitoring data and simulation forecast data in real time, and using the text embedding model to perform cluster analysis on the current basic monitoring data and simulation forecast data, as well as the historical monitoring data and historical simulation data sets, the target clustering categories corresponding to the current basic monitoring data and simulation forecast data can be accurately obtained, and the corresponding warning release message can be determined according to the target clustering category, and used as the initial warning release message; the historical warning release message and the initial warning release message are processed by the large language model, so as to output the target warning release message for the later stage for decision-making assistance. By combining the text embedding model and the large language model, the data at different stages are processed respectively, and the efficiency of knowledge acquisition and knowledge utilization is further improved in the face of the highly professional waterlogging warning release work. By acquiring and processing the current basic monitoring data and simulation forecast data in real time, automatic warning release and immediate and rapid warning release can be realized; the function of early forecast and warning can enable emergency response work to be advanced, which can reduce the losses caused by disasters. And compared with the existing technology, this scheme has the advantages of moderate cost, good effect and low technical difficulty. In addition, by optimizing the historical warning release message, the obsolescence problem of the existing scheme can be reviewed in real time. Through manual experience analysis, analysis and review by experts, it is determined what kind of warning message should be issued at the current time in history, and the latest warning messages and emergency response plans can be used in the face of climate change and land use and cover changes. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0041] Figure 1 A schematic diagram of an application environment of the waterlogging warning information processing method of the present invention;
[0042] Figure 2 It is a flow chart of the waterlogging warning information processing method of the present invention;
[0043] Figure 3 A flowchart of a method for processing a text embedding model of the present invention;
[0044] Figure 4 A flow chart of a method for establishing a collection of historical monitoring data and early warning release messages according to the present invention;
[0045] Figure 5 is a schematic diagram of a waterlogging warning information processing system of the present invention;
[0046] Figure 6 Schematic diagram of an electronic device of the present invention. DETAILED DESCRIPTION
[0047] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0048] In the relevant technology, traditional waterlogging warning messages are generally released according to agreed specifications; however, when the relevant specification information is outdated or too little information is available, it is easy to cause the defect of poor emergency response efficiency after the warning message is released.
[0049] In order to solve the technical defects existing in the related art, in this exemplary embodiment, a method for processing waterlogging warning information is provided. Specifically, the method can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network 101. The data storage system 103 can store the data that the server 104 needs to process. The data storage system 103 can be integrated on the server 104, or it can be placed on the cloud or other network servers. Among them, the terminal 102 can be but is not limited to various personal computers, laptops, smart phones, tablets and portable wearable devices. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The head-mounted device can be a virtual reality (VR) device, an augmented reality (AR) device, smart glasses, etc. The server 104 can be an independent physical server, or it can be a server cluster or distributed system composed of multiple physical servers, or it can be a cloud server that provides cloud computing services. Among them, the user information collected by the terminal 102 and the user information processed by the server 104 can be information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.
[0050] For example, a user may initiate a data processing request for waterlogging warning information on the terminal 102, and the server 104 may receive and execute the data processing request initiated by the terminal, thereby obtaining the current basic monitoring data and the current simulation forecast data, and storing them in the data storage system 103. The server may obtain the current basic monitoring data, the current simulation forecast data, and related historical data from the data storage system, and call the waterlogging warning information processing method to parse the obtained data to obtain a target warning release message for assisting decision-making; and then send these target warning release message terminals to assist users in making decisions.
[0051] In an exemplary embodiment, Figure 2 As shown, a method for processing waterlogging warning information is provided, and the method is applied to Figure 1 The server in is used as an example to illustrate, including the following steps S11 to S13. Specifically,
[0052] In step S11, the current basic monitoring data and the current simulation forecast data are obtained; wherein the current basic monitoring data are used to describe the real-time water situation data currently collected; and the current simulation forecast data are used to describe the simulated water situation data corresponding to a preset time period in the future after the current basic monitoring data is obtained.
[0053] Exemplarily, the above-mentioned current basic monitoring data may be real-time water condition data transmitted back by water condition equipment at the designated monitoring area through the Internet of Things, including but not limited to water condition data such as rainfall, water level, and water depth. Among them, the current simulation forecast data may be forecast data for a period of time from now to the future obtained by simulation of basic or non-basic models, such as water condition data such as rainfall, water level, and water depth in the next three days or a week. Among them, the basic monitoring data and the simulation forecast data may correspond to the monitoring data of geographic information data, and the structure of geographic information data may be a one-dimensional, two-dimensional, or three-dimensional structure. For example, the simulation forecast data may be forecast rainfall data provided by a meteorological unit, forecast waterlogging data simulated by a model, and so on.
[0054] Exemplarily, the format and content of the real-time current basic monitoring data and the current simulation forecast data can be expressed as:
[0055]
[0056] in, It is a collection of current basic monitoring data and current simulation forecast data; Represents the nth type of data, which is one of the current basic monitoring data and the current simulation forecast data.
[0057] In step S12, the text embedding model is used to cluster the current basic monitoring data, the current simulation forecast data, and the historical monitoring data and the historical simulation forecast data to determine the target clustering categories corresponding to the current basic monitoring data and the current simulation forecast data, and the initial warning release message corresponding to the target clustering categories; wherein the initial warning release message is determined by the historical monitoring data and the warning release message set corresponding to the same target clustering category.
[0058] In this exemplary embodiment, reference Figure 3 As shown, the above step S12 may include:
[0059] Step S121: collect the current basic monitoring data and the current simulation forecast data The historical monitoring data and historical simulation forecast data set X are simultaneously input into the text embedding model for clustering analysis to determine the clustering categories of each subset in the set, expressed as Among them, C is the cluster category set; Represents the current basic monitoring data and the current simulation forecast data set Cluster category; c m Represents the clustering categories of each subset in the historical monitoring data and historical simulation forecast data set X;
[0060] Step S122: based on the current basic monitoring data and the current simulation forecast data set The mapping relationship between the historical monitoring data and the historical simulation forecast data set X is used to determine the current basic monitoring data and the current simulation forecast data set Clustering categories
[0061] Step S123, based on the clustering category and historical monitoring data subsets (c1, c2, ..., c m )’s intersection determines the same cluster category c′;
[0062] Step S124, determining a subset x' of historical warning data with the same cluster category c' corresponding to the same cluster category according to the mapping relationship between the historical monitoring data and the historical simulation forecast data set X and the cluster category set C;
[0063] Step S125, based on the correspondence between the historical monitoring data set and the optimized warning release message set, determine the search result y′ in the optimized warning release message set with the same clustering category in the historical monitoring data and warning release message set M as the initial warning release message.
[0064] Exemplarily, in step S121, the text embedding model may be an unsupervised clustering model, such as a K-means model, a self-organizing map model, a neighbor propagation algorithm model, a hierarchical clustering model, a mean shift algorithm model, a DBSCAN model, a binary K-means model, etc.
[0065] Exemplarily, the current basic monitoring data and the current simulation forecast data are combined The historical monitoring data and the historical simulation forecast data set X are simultaneously input into the text embedding model for cluster analysis. The text embedding model is input for cluster analysis, including: randomly selecting k clusters as the initial cluster center P0;
[0066] For the current basic monitoring data and the current simulation forecast data set For each data point in the historical monitoring data and historical simulation forecast data set X, calculate its distance from each cluster center and assign it to the cluster center category C with the closest distance k , is the set of data points of the kth cluster;
[0067] For k clusters, recalculate each cluster center; wherein the updated cluster center is determined based on the mean of all data points in the cluster;
[0068] Iterate based on the updated cluster centers until the preset termination condition is met to complete clustering.
[0069] For example, taking the K-means model as an example, the current basic monitoring data and the simulated forecast data set When the text embedding model is input into the historical monitoring data set X for clustering analysis, the following steps can be included:
[0070] Step 1: At the beginning of the K-means model algorithm, k clusters are randomly selected as the initial cluster centers P0. The format and content of the initial cluster center set of k clusters are the same as is consistent with X. Among them, the initial cluster center P0 is as follows:
[0071]
[0072] Among them, P0 is the initial cluster center set; p kn Represents the nth type of data of the initial cluster center of the kth cluster.
[0073] Step 2: For the dataset For each data point in the dataset X, calculate its distance from each cluster center and assign it to the cluster center category C with the closest distance k , is the set of data points of the kth cluster.
[0074] Step 3: For k clusters, recalculate each cluster center. The new cluster center is the mean of all data points in the cluster, and the calculation formula is as follows:
[0075]
[0076] Among them, P t is the set of cluster centers of the tth iteration; The new center of the k-th cluster of the n-th data at the t-th iteration; x n Is a collection and the set X that belongs to C k Collect the nth type of data; |C k | is the number of data points in the kth cluster.
[0077] Step 4: Jump to step 2 to start iteratively repeating the allocation and updating steps until the termination condition is met. The termination condition includes that the cluster center no longer changes significantly and the maximum number of iterations is reached.
[0078] Finally, the clustering results can be used to compare the categories to which each sub-set in the set belongs. The formula may include:
[0079]
[0080] Among them, C is the cluster category set; Representative basic monitoring data and simulation forecast data set Cluster category; c m Represents the clustering categories of each subset in the historical monitoring data X. Among them, the clustering category may be 2 or more categories, so the categories in each subset in the clustering category set may be the same or different; the basic monitoring data and simulation forecast data set There is a corresponding relationship between the historical monitoring data and the historical simulation forecast data set X and the cluster category set C, for example, it belongs to a bijective relationship.
[0081]
[0082] Exemplarily, the optimization warning release message of the same cluster category is found through the cluster category set C as the search result. The calculation process formula can be expressed as:
[0083]
[0084] Among them, the basic monitoring data and simulation forecast data set are considered The bijective relationship between the cluster category set C can be used to find the basic monitoring data and the simulated forecast data set. Clustering categories pass and subsets (c1,c2,…,c m ) The intersection finds the same cluster category c′.
[0085] Since there is a correspondence between the historical monitoring data and the historical simulation forecast data set X and the cluster category set C, such as a bijective relationship, it is possible to find a historical data subset x' corresponding to the same cluster category c'. Since the historical monitoring data and the optimized warning release message set are bijective, the optimized warning release message set of the same cluster category can be found as the search result y' through the historical monitoring simulation data and the warning release message set M.
[0086] In this exemplary embodiment, the above historical monitoring data and warning release message set can be pre-established based on historical basic monitoring data and historical simulation forecast data. Figure 4 As shown, before clustering the current basic monitoring data, the current simulation forecast data, and the historical monitoring data and historical simulation forecast data set, the method further includes:
[0087] Step S21, constructing a set of historical monitoring data and historical simulation forecast data according to the historical basic monitoring data and historical simulation forecast data corresponding to each historical time point;
[0088] Step S22, obtaining historical warning release messages corresponding to the historical basic monitoring data and historical simulation forecast data at each historical time point, optimizing the historical warning release messages to obtain corresponding optimized warning release messages, and constructing an optimized warning message set;
[0089] Step S23, combining and processing the historical monitoring data, the historical simulation forecast data set, and the optimized early warning release message set based on historical time points to establish the historical monitoring data and early warning release message set.
[0090] For example, in the above step S21, specifically, the historical basic monitoring data and historical simulation forecast data corresponding to multiple different historical time points can be used to form a set of historical monitoring data and historical simulation forecast data, and the format and content of the single data in the set must be consistent with the basic monitoring data and simulation forecast data obtained in real time. Specifically, it can be expressed as:
[0091] X={(x 11 , x 12 , …, x 1n ), (x 21 , x 22 , …, x 2n ),…,(x m1 , x m2 , …, x mn )} (6)
[0092] Where X is the set of historical monitoring data and historical simulation forecast data; () represents the subset at the mth moment; x mn Represents the nth type of data at the mth moment.
[0093] Exemplarily, in the above-mentioned step S22, the historical warning release messages are optimized to obtain corresponding optimized warning release messages, including: establishing manual analysis tasks for the historical warning release messages at each historical time point, performing data analysis on the historical basic monitoring data and historical simulation forecast data corresponding to each historical time point, and optimizing the historical warning release messages according to the data analysis results to determine the corresponding optimized warning release messages.
[0094] Specifically, by establishing and processing a manual analysis task, the historical flooding situation, rainfall data, and water data are analyzed, analyzed, and reviewed by experts through manual experience, to obtain the most correct warning release message that should be issued at the current time in history. In addition, other data other than historical basic monitoring data can also be referenced during data analysis, such as location information, evacuation notification information, information notification method, number of disaster-stricken people, number of staff, and other data. For example, if the rainfall is expected to reach 150 mm in the next 3 hours, a street or community will be issued to evacuate people and vehicles; if it is monitored that the water accumulation in a tunnel under a certain section of road has reached 50 cm, a tunnel under a certain section of road will be closed. Finally, a set of optimized warning release messages at different historical time points is formed, and the formula is expressed as follows:
[0095] Y={(y 11 ,y 12 ,…),(y 21 ,y 22 ,…),…,(y m1 ,y m2 , …)} (7)
[0096] Among them, Y is the optimized warning release message set; y m1 Represents the first message at the mth moment. The number of subsets at each moment may vary depending on the time and on-site conditions. It may be an empty set, or there may be one or more sets.
[0097] For example, in the above step S23, based on the historical basic monitoring data, the historical simulation forecast data, and the optimized warning release message corresponding to the historical time point, a data set of historical monitoring data and warning release message is established; it is expressed as:
[0098] M={(x1:y1),(x2:y2),…,(x m :y m )} (8)
[0099] Among them, M is the set of historical monitoring data and warning release messages; x m Represents the historical data subset at the mth moment (x m1 ,x m2 ,…,x mn );y m Represents the optimized warning release message subset (y m1 ,y m2 ,…). The historical monitoring data and optimization warning release message collection can be a bijective relationship
[0100] The optimized warning release message set includes a historical monitoring data and historical simulation forecast data set X and a historical monitoring data and warning release message set M.
[0101] This solution can obtain the best warning release information corresponding to the historical monitoring data by optimizing the existing historical warning release messages, so as to facilitate the subsequent accurate warning messages for the current monitoring data.
[0102] In step S13, the historical warning release message and the initial warning release message are input into the trained large language model, and the target warning release message output by the large language model for assisting decision making is obtained.
[0103] Exemplarily, the historical warning release message is the historical warning release message that has been released before the current moment, and the corresponding optimized warning release message, as the warning release message of the previous stage. The retrieval result output by the text embedding model, that is, the initial warning release message, and the historical warning release message can be input into the large language model to infer the target warning release message of the later stage for auxiliary decision-making. The large language model is used to match the initial warning release message with the historical warning release message, obtain the optimized warning release message that matches the current basic monitoring data and simulated forecast data, and output it as the target warning release message.
[0104] Exemplarily, the large language model can be trained in advance, including a pre-training step and a large model fine-tuning step. Among them, the historical monitoring data set corresponding to each historical time point, the historical warning release message, and the corresponding optimized warning release message set can be input into the preset large language model, and the large language model is fine-tuned with the optimized warning release message input by the large language model as the optimization target. Among them, the training method of the large language model can be implemented based on the existing technology, and the present application does not specifically limit the type of large language model and the training process. For example, the above-mentioned large language model includes but is not limited to GPT, LaMDA, LLaMA, Gemini, Wenxin Yiyan, Baichuan, Xinghuo, Tongyi Qianwen, Hunyuan and other deep learning models, which can be used to generate natural language text or understand the meaning of language text.
[0105] The warning information processing method provided by the present invention couples the text embedding model with the large language model, and further improves the efficiency of knowledge acquisition and knowledge utilization in the highly professional work of issuing waterlogging warnings. By acquiring and processing the current basic monitoring data and simulated forecast data in real time, it is possible to realize automatic warning release and immediate and rapid warning release; the function of early forecast and warning enables emergency response work to be advanced, which can reduce the losses caused by disasters. And compared with the existing technology, this solution has the advantages of moderate cost, good effect and low technical difficulty. In addition, through the optimization of historical warning release messages, the obsolescence problem of existing solutions can be reviewed in real time. Through the manual experience analysis, analysis and review of experts, it is known what kind of warning message should be released in history, and the latest warning release message and emergency response plan can be used in the face of climate change and land use and cover changes.
[0106] Based on the same inventive concept, the embodiment of the present application also provides a system for implementing the above-mentioned waterlogging warning information processing method. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more waterlogging warning information processing system embodiments provided below can refer to the limitations of the waterlogging warning information processing method above, and will not be repeated here.
[0107] In one embodiment, a waterlogging warning information processing system is provided, referring to Figure 5 As shown, the system 50 includes:
[0108] The data acquisition module 501 is used to acquire current basic monitoring data and current simulation forecast data; wherein the current basic monitoring data is used to describe the real-time water condition data currently collected; and the current simulation forecast data is used to describe the simulated water condition data corresponding to a preset time period in the future after the current basic monitoring data is acquired;
[0109] The text embedding model processing module 502 is used to cluster the current basic monitoring data, the current simulation forecast data, and the historical monitoring data and the historical simulation forecast data set using the text embedding model, and determine the target clustering category corresponding to the current basic monitoring data and the current simulation forecast data, and the initial warning release message corresponding to the target clustering category; wherein the initial warning release message is determined by the same target clustering category corresponding to the historical monitoring data and the warning release message set;
[0110] The large language model processing module 503 is used to input the historical warning release message and the initial warning release message into the trained large language model, and obtain the target warning release message output by the large language model for auxiliary decision-making.
[0111] Each module in the above-mentioned waterlogging warning information processing system can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.
[0112] In one embodiment, an electronic device is provided, whose structure diagram can be as follows: Figure 6 As shown. The device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface, the display unit and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a multi-hop proximity service communication method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.
[0113] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0114] In one embodiment, an electronic device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.
[0115] It should be noted that, as another aspect, the present application also provides a storage medium, which can be included in an electronic device; or it can exist independently without being installed in the electronic device. The above storage medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the method described in the above embodiment. For example, the electronic device can implement the following method: Figure 2 The steps shown.
[0116] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.
[0117] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above-mentioned method embodiments when executed by a processor. The computer program can be loaded onto a computer or other programmable data processing device, so that a series of operating steps are performed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the above-mentioned method embodiments.
[0118] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to an exemplary embodiment of the present invention, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.
[0119] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for processing waterlogging warning information, characterized in that: The method comprises: Acquire current basic monitoring data and current simulation forecast data; wherein the current basic monitoring data is used to describe the real-time water condition data currently collected; and the current simulation forecast data is used to describe the simulated water condition data corresponding to a preset time period in the future after the current basic monitoring data is acquired; The text embedding model is used to cluster the current basic monitoring data, the current simulation forecast data, and the historical monitoring data and the historical simulation forecast data set, and determine the target clustering category corresponding to the current basic monitoring data and the current simulation forecast data, and the initial warning release message corresponding to the target clustering category; wherein the initial warning release message is determined by the same target clustering category corresponding to the historical monitoring data and the warning release message set; The historical warning release message and the initial warning release message are input into the trained large language model, and the target warning release message output by the large language model for assisting decision making is obtained.
2. The method for processing waterlogging warning information according to claim 1, characterized in that: Before clustering the current basic monitoring data, the current simulation forecast data, and the historical monitoring data and historical simulation forecast data set, the method further includes: According to the historical basic monitoring data and historical simulation forecast data corresponding to each historical time point, a collection of historical monitoring data and historical simulation forecast data is constructed; Obtaining historical warning release messages corresponding to the historical basic monitoring data and historical simulation forecast data at each historical time point, optimizing the historical warning release messages to obtain corresponding optimized warning release messages, and constructing an optimized warning message set; The historical monitoring data, the historical simulation forecast data set, and the optimized early warning release message set are combined and processed based on historical time points to establish the historical monitoring data and early warning release message set.
3. The method for processing waterlogging warning information according to claim 2, characterized in that: The optimizing process of the historical warning release message to obtain the corresponding optimized warning release message includes: Establish manual analysis tasks for historical warning release messages at each historical time point, perform data analysis on the historical basic monitoring data and historical simulation forecast data corresponding to each historical time point, optimize the historical warning release messages based on the data analysis results, and determine the corresponding optimized warning release messages.
4. The method for processing waterlogging warning information according to claim 2 or 3, characterized in that: The optimization warning release message set includes optimization warning release messages at multiple different historical time points, expressed as: And={(and 11 ,and 12 ,…),(and 21 ,and 22 ,…),…,(and m1 ,and m2 ,…)} Among them, Y is the set of optimized warning release messages; m1 Represents the first message at the mth moment, and the number of subsets at each moment is different.
5. The method for processing waterlogging warning information according to claim 2, characterized in that: The optimized warning release message set includes: a historical monitoring data and historical simulation forecast data set X, and a historical monitoring data and warning release message set M; Where X = {(x 11 ,x 12 ,…,x 1n ),(x 21 ,x 22 ,…,x 2n ),…,(x m1 ,x m2 ,…,x mn )}; where () represents the subset at the mth moment; x mn Represents the nth type of data at the mth moment; M={(x1:y1),(x2:y2),…,(x m :y m )}; x m Represents the historical monitoring data subset at the mth moment (x m1 ,x m2 ,…,x mn );y m Represents the optimized warning release message subset at the mth moment (y m1 ,y m2 ,…).
6. The method for processing waterlogging warning information according to claim 1, characterized in that: The method of clustering the basic monitoring data, the simulated forecast data, and the historical monitoring data and the historical simulated forecast data set using the text embedding model, and determining the target clustering categories corresponding to the basic monitoring data and the simulated forecast data, and the initial warning release message corresponding to the target clustering category, includes: Combine the current basic monitoring data with the current simulation forecast data The historical monitoring data and historical simulation forecast data set X are simultaneously input into the text embedding model for clustering analysis to determine the clustering categories of each subset in the set, expressed as Among them, C is the cluster category set; Represents the current basic monitoring data and the current simulation forecast data set Cluster category; c m Represents the clustering categories of each subset in the historical monitoring data and historical simulation forecast data set X; Based on the current basic monitoring data and current simulation forecast data set The mapping relationship between the historical monitoring data and the historical simulation forecast data set X is used to determine the current basic monitoring data and the current simulation forecast data set Clustering categories According to clustering categories and historical monitoring data subsets (c1, c2, ..., c m )’s intersection determines the same cluster category c′; According to the mapping relationship between the historical monitoring data and the historical simulation forecast data set X and the cluster category set C, the historical warning data subset x′ with the same cluster category c′ is determined; Based on the correspondence between the historical monitoring data set and the optimized warning release message set, a search result y′ in the optimized warning release message set with the same clustering category is determined in the historical monitoring data and warning release message set M as the initial warning release message.
7. The method for processing waterlogging warning information according to claim 6, characterized in that: The current basic monitoring data and the current simulation forecast data are combined The historical monitoring data and the historical simulation forecast data set X are simultaneously input into the text embedding model for cluster analysis. The text embedding model is input for cluster analysis, including: Randomly select k clusters as the initial cluster center P0; For the current basic monitoring data and the current simulation forecast data set For each data point in the historical monitoring data and historical simulation forecast data set X, calculate its distance from each cluster center and assign it to the cluster center category C with the closest distance k , is the set of data points of the kth cluster; For k clusters, recalculate each cluster center; wherein the updated cluster center is determined based on the mean of all data points in the cluster; Iterate based on the updated cluster centers until the preset termination condition is met to complete clustering.
8. The method for processing waterlogging warning information according to claim 1, characterized in that: The current basic monitoring data includes: at least one of rainfall, water level and water depth data; The current simulation forecast data includes: at least one of the rainfall, water level and water depth data estimated based on the basic monitoring data.
9. A waterlogging warning information processing system, characterized in that: The system comprises: A data acquisition module, used to acquire current basic monitoring data and current simulation forecast data; wherein the current basic monitoring data is used to describe the real-time water condition data currently collected; and the current simulation forecast data is used to describe the simulated water condition data corresponding to a preset time period in the future after the current basic monitoring data is acquired; A text embedding model processing module, used to cluster the current basic monitoring data, the current simulation forecast data, and the historical monitoring data and historical simulation forecast data sets using a text embedding model, and determine the target clustering categories corresponding to the current basic monitoring data and the current simulation forecast data, and the initial warning release message corresponding to the target clustering category; wherein the initial warning release message is determined by the same target clustering category corresponding to the historical monitoring data and the warning release message set; The large language model processing module is used to input the historical warning release message and the initial warning release message into the trained large language model, and obtain the target warning release message output by the large language model for auxiliary decision-making.
10. An electronic device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; The processor is configured to execute the waterlogging warning information processing method according to any one of claims 1 to 8 by executing the executable instructions.
Citation Information
Patent Citations
Field monitoring and numerical simulation combined mine disaster forecasting and pre-warning method
CN108510112A
Fan fault early warning method and system
CN111209934A
Power grid thunder and lightning early warning method, device, recording medium and system
CN113705693A
Urban inland inundation early warning method, device and equipment and storage medium
CN114662894A
Early warning method and system for geological disaster monitoring
CN115034600A
