Digital twin water conservancy monitoring method and system based on artificial intelligence

Abstract

The invention discloses a digital twinborn water conservancy monitoring method and system based on artificial intelligence, and relates to the technical field of water conservancy projects, water conservancy object information is determined, and a digital twinborn model is constructed according to the water conservancy object information; binding a water conservancy calculation model corresponding to the digital twinborn model, and obtaining a plurality of calculation environments based on divergence of the water conservancy calculation model; water conservancy data are collected in real time based on water conservancy object information, the digital twin model is updated in real time, water quality data of a water conservancy object are calculated according to the water conservancy calculation model, and target water quality development prediction information is obtained. According to the method, the water quality development prediction information corresponding to the maximum number of the ensemble twinning water conservancy calculation models is used as the target water quality development prediction information, the multiple twinning water conservancy calculation models are mutually independent, long-term calculation information of the water quality information can be provided, and subsequent water quality maintenance by workers is facilitated.

Description

Technical Field

[0001] The present invention relates to the technical field of water conservancy engineering, and in particular to a digital twin water conservancy monitoring method and system based on artificial intelligence. Background Art

[0002] With the increasing scarcity of global water resources and the frequent occurrence of water quality problems, water resource management faces unprecedented challenges; with the continuous development and improvement of artificial intelligence technology, water quality prediction still relies on a single or a small number of models, which lack sufficient data for comparison, resulting in low prediction accuracy; insufficient accuracy can lead to incorrect water quality management decisions, which in turn affects the protection and utilization of water resources; in addition, during data collection and transmission, water quality information collected at monitoring points may not be processed and utilized in a timely manner; information delays may lead to delayed responses to water quality changes and miss the best time for processing. Summary of the invention

[0003] The purpose of the present invention is to provide a digital twin water conservancy monitoring method and system based on artificial intelligence to solve the shortcomings of the background technology.

[0004] In order to achieve the above object, the present invention provides the following technical solution: a digital twin water conservancy monitoring method based on artificial intelligence, comprising the following steps: Determine water conservancy object information, where the water conservancy object information includes water conservancy monitoring scope and monitoring parameter type, and build a digital twin model based on the water conservancy object information; The corresponding digital twin model is bound to the water conservancy calculation model, and multiple calculation environments are obtained based on the divergence of the water conservancy calculation model; Based on the information of water conservancy objects, water conservancy data is collected in real time and the digital twin model is updated in real time. The water quality data of the water conservancy objects is calculated according to the water conservancy calculation model to obtain the target water quality development prediction information.

[0005] In a preferred embodiment, the step of constructing a digital twin model based on water conservancy object information includes: Define the water conservancy monitoring scope, determine the types of monitoring parameters that need to be collected within the water conservancy monitoring scope, and use the water conservancy monitoring scope and monitoring parameter types as water conservancy object information; Set up multiple monitoring points within the water conservancy monitoring range corresponding to the monitoring parameter types; A water conservancy model is constructed according to the water conservancy monitoring scope, corresponding monitoring points are marked in the water conservancy model to obtain marking points, and corresponding monitoring scopes are divided according to corresponding marking points on the water conservancy model; The corresponding cloud server is configured for the corresponding marking point, and the cloud server is connected to the corresponding monitoring point for communication to obtain a digital twin model.

[0006] In a preferred embodiment, the corresponding digital twin model is bound to a water conservancy calculation model, and the step of obtaining multiple calculation environments based on the water conservancy calculation model comprises: Obtain the historical water quality information of each monitoring point within the water conservancy monitoring range, and connect the monitoring points according to the water flow direction within the water conservancy monitoring range to obtain a monitoring point chain; Corresponding learning models are configured between adjacent monitoring points in the monitoring point chain, and the learning models corresponding to the monitoring points are trained through the historical water quality information of the adjacent monitoring points to obtain the target learning model; Based on the monitoring point chain, the target learning models corresponding to the adjacent monitoring points are arranged and connected to obtain the water conservancy estimation model; The water conservancy calculation model is diverged to obtain multiple calculation environments.

[0007] In a preferred embodiment, the step of diverging the water conservancy estimation model to obtain multiple estimation environments includes: A plurality of divergence boxes are set corresponding to the water conservancy estimation model, and the number of the divergence boxes is the same as the number of the target learning models in the water conservancy estimation model, wherein the divergence box includes a plurality of guide replication points; Set a corresponding number of virtual spaces according to the number of divergent frames; The water conservancy calculation model is copied incrementally through the divergence frame to obtain multiple twin water conservancy calculation models, and the binding between the water conservancy calculation model and the digital twin model is retained; Multiple twin water conservancy calculation models are stored one-to-one in the corresponding virtual space through a divergent frame to obtain multiple calculation environments.

[0008] In a preferred embodiment, the step of obtaining multiple twin hydraulic calculation models by decrementally replicating the hydraulic calculation model through a divergent frame includes: Arrange multiple divergent frames, enable multiple guide replication points in the first divergent frame in the arrangement order, the number of the multiple guide replication points is the same as the number of target learning models in the water conservancy estimation model, and reduce the number of enabled guide replication points by one for the divergent frames after the arrangement order; The target learning model in the water conservancy estimation model is sequentially and descendingly replicated by the guided replication points enabled in the multiple divergent boxes to obtain multiple twin water conservancy estimation models; Based on the guided replication points, perception channels are established between the water conservancy calculation model and multiple twin water conservancy calculation models.

[0009] In a preferred embodiment, the step of establishing perception channels between the hydraulic estimation model and multiple twin hydraulic estimation models based on the guide replication points includes: The guided replication point includes a sensing space, a plurality of sensing points in the sensing space, and water quality information configured at the sensing points; Based on the guided replication point, a transmission channel is established between the hydraulic calculation model and the first upstream target learning model in multiple twin hydraulic calculation models, the sensing point is connected to the target learning model in the hydraulic calculation model that is the same as the first upstream target learning model in the twin hydraulic calculation model, and the sensing attraction point and the water quality information configured by the sensing attraction point are set in the target learning model that is the same as the first upstream target learning model in the twin hydraulic calculation model. The sensing point will move toward the sensing attraction point with the same water quality information, and the sensing space is communicated and connected with the first upstream target learning model in multiple twin hydraulic calculation models to obtain a perception channel between the hydraulic calculation model and the replicated target learning models in multiple twin hydraulic calculation models.

[0010] In a preferred embodiment, the step of collecting water conservancy data in real time based on water conservancy object information and updating the digital twin model in real time, and calculating the water quality data of the water conservancy object according to the water conservancy calculation model to obtain target water quality development prediction information includes: Collect water quality information in real time based on monitoring points and update the water quality information to the digital twin model; The water quality information in the corresponding digital twin model is obtained through the hydraulic calculation model. The hydraulic calculation model derives the corresponding sensing attraction point according to the water quality information of the monitoring point, and attracts the sensing point corresponding to the water quality information through the transmission channel between the hydraulic calculation model and the first target learning model upstream of the multiple twin hydraulic calculation models; When the sensing point is out of the sensing space, the first target learning model upstream of the twin water conservancy calculation model will calculate the water quality information corresponding to the sensing points that are missing in the sensing space, and will collect the twin water conservancy calculation models with overlapping water quality development prediction information, and will use the corresponding water quality development prediction information of the twin water conservancy calculation models with the largest number of collected twin water conservancy calculation models as the target water quality development prediction information.

[0011] The present invention also provides a digital twin water conservancy monitoring system based on artificial intelligence, including: A construction module is used to determine water conservancy object information, wherein the water conservancy object information includes the water conservancy monitoring scope and the monitoring parameter type, and to construct a digital twin model according to the water conservancy object information; The setting module is connected with the construction module and is used to bind the water conservancy calculation model to the corresponding digital twin model, and obtain multiple calculation environments based on the water conservancy calculation model; The calculation module is connected to the setting module, and is used to collect water conservancy data in real time based on the water conservancy object information and update the digital twin model in real time. The water quality data of the water conservancy object is calculated according to the water conservancy calculation model to obtain the target water quality development prediction information.

[0012] In the above technical solution, the technical effects and advantages provided by the present invention are: 1. The present invention can aggregate twin water conservancy estimation models with overlapping water quality development prediction information, which means that the twin water conservancy estimation models with repeated estimation results are more accurate. Then, the corresponding water quality development prediction information of the largest number of the aggregated twin water conservancy estimation models is used as the target water quality development prediction information. Multiple twin water conservancy estimation models are independent of each other, which can improve the long-term estimation information of water quality information and facilitate the staff to perform subsequent water quality maintenance; 2. The present invention can also be quickly acquired by the twin water conservancy calculation model, and can adapt to the high efficiency of the transmission of water quality information collected by monitoring points. When the water quality information changes in a short period of time, it can be quickly acquired and changed by the twin water conservancy calculation model, thereby ensuring the timeliness and adaptability of the twin water conservancy calculation model's calculated data. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0014] Figure 1 The present invention is a flow chart of the method.

[0015] Figure 2 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0017] Example 1, please refer to Figure 1 As shown, the digital twin water conservancy monitoring method based on artificial intelligence described in this embodiment includes the following steps: S1. Determine water conservancy object information, where the water conservancy object information includes water conservancy monitoring scope and monitoring parameter type, and build a digital twin model based on the water conservancy object information; S2, the corresponding digital twin model is bound to the water conservancy calculation model, and multiple calculation environments are obtained based on the divergence of the water conservancy calculation model; S3. Collect water conservancy data in real time based on water conservancy object information and update the digital twin model in real time. Calculate the water quality data of the water conservancy object according to the water conservancy calculation model to obtain target water quality development prediction information.

[0018] As described in the above steps S1-S3, the twin water conservancy calculation models with overlapping water quality development prediction information can be aggregated, which means that the twin water conservancy calculation models with repeated calculation results are more accurate. Then, the water quality development prediction information corresponding to the largest number of twin water conservancy calculation models is taken as the target water quality development prediction information. Multiple twin water conservancy calculation models are independent of each other, which can improve the long-term calculation information of water quality information and facilitate the staff to carry out subsequent water quality maintenance; they can also be quickly acquired by the twin water conservancy calculation model and can adapt to the high efficiency of the transmission of water quality information collected by the monitoring points. When the water quality information changes in a short period of time, it can be quickly acquired and changed by the twin water conservancy calculation model to ensure the timeliness and adaptability of the twin water conservancy calculation model data.

[0019] In one embodiment, the step S1 of constructing a digital twin model according to water conservancy object information includes: S11, delineate the water conservancy monitoring scope, determine the types of monitoring parameters that need to be collected within the water conservancy monitoring scope, and use the water conservancy monitoring scope and monitoring parameter types as water conservancy object information; S12. Set multiple monitoring points corresponding to the monitoring parameter types within the water conservancy monitoring range; S13, constructing a water conservancy model according to the water conservancy monitoring range, marking corresponding monitoring points in the water conservancy model to obtain marking points, and dividing corresponding monitoring ranges on the water conservancy model according to corresponding marking points; S14, configuring a corresponding cloud server for the corresponding marking point, and communicating and connecting the cloud server with the corresponding monitoring point to obtain a digital twin model; As described in the above steps S11-S14, the scope of water conservancy monitoring is determined, for example, it can be a river area, and the type of monitoring parameters that need to be collected is determined corresponding to the water conservancy monitoring scope, such as various parameter information on water quality, which is used to monitor the safety and development of water quality. The water conservancy monitoring scope and the monitoring parameter type are used as water conservancy object information, and multiple monitoring points are set within the water conservancy monitoring scope corresponding to the monitoring parameter type. The monitoring points are monitoring instruments set for the corresponding monitoring parameter types, which are used to collect water quality information within the water conservancy monitoring scope. A corresponding three-dimensional model is constructed as a water conservancy model according to the water conservancy monitoring scope, and the positions of the corresponding actual monitoring points are marked in the water conservancy model to obtain marked points. The corresponding monitoring range is divided according to the marked points on the water conservancy model. The monitoring range is the range for monitoring water quality by the marked points. The corresponding cloud server is configured for the corresponding marked points, and the cloud server is connected to the corresponding monitoring points for communication. This can facilitate the real-time transmission of water quality information collected by subsequent monitoring points to the cloud server, and update the digital twin model in real time.

[0020] In one embodiment, the corresponding digital twin model is bound to a water conservancy calculation model, and step S2 of obtaining multiple calculation environments based on the water conservancy calculation model diverges includes: S21, obtaining historical water quality information of each monitoring point within the water conservancy monitoring range, and connecting the monitoring points according to the water flow direction within the water conservancy monitoring range to obtain a monitoring point chain; S22, configuring corresponding learning models between adjacent monitoring points in the monitoring point chain, and training the learning models corresponding to the monitoring points through historical water quality information of the adjacent monitoring points to obtain a target learning model; S23, arranging and connecting the target learning models corresponding to the adjacent monitoring points based on the monitoring point chain to obtain a water conservancy estimation model; S24, diverging the water conservancy calculation model to obtain multiple calculation environments; As described in the above steps S21-S24, multiple monitoring points are set within the water conservancy monitoring range, and the historical water quality information of each monitoring point within the water conservancy monitoring range is obtained. The monitoring points are connected according to the water flow direction within the water conservancy monitoring range to obtain a monitoring point chain. In the water conservancy monitoring range, the water flow direction will pass through the monitoring points in a front-to-back order, so the water quality information is affected by the upstream to the downstream water quality. Therefore, corresponding learning models are configured between adjacent monitoring points in the monitoring point chain, and the learning model corresponding to the monitoring point is trained with the historical water quality information of the adjacent monitoring points to obtain a target learning model. The training method of the learning model is: include the historical water quality information of the adjacent monitoring points in the current monitoring point The historical water quality information is divided into a training set and a test set, wherein the historical water quality information in the training set and the test set are the corresponding water quality information of the current monitoring point and the water quality information of the next monitoring point of the current monitoring point. The water quality information includes water flow rate, ambient temperature and pollutant parameters of water quality. The historical water quality information of adjacent monitoring points is preprocessed. The preprocessing here is data cleaning, and then the learning model is trained with the training set after preprocessing, and then the trained learning model is tested with the test set until the learning model meets the inference requirements, and the target learning model is obtained. It is used to train the learning model to analyze the changes in water quality information between adjacent monitoring points, and then arrange and connect the target learning models corresponding to the adjacent monitoring points according to the monitoring point chain to obtain a water conservancy calculation model. The arrangement here arranges and links the target learning models according to the order of water flow of the monitoring points, and uses the output value of the previous target learning model as the input value of the next target learning model. The target learning model here inputs the water quality information of the current monitoring point, and then can calculate the future water quality information of the next monitoring point. The calculated water quality information can provide effective data for water conservancy management for staff, and better control water pollution. The water conservancy calculation model is diverged to obtain multiple calculation environments, which can segment the water quality information collected by the monitoring points within the water conservancy monitoring range for analysis, thereby improving the accuracy of long-term calculation of water quality information within the water conservancy monitoring range, and being more in line with actual conditions.

[0021] In one embodiment, the step S24 of diverging the water conservancy estimation model to obtain multiple estimation environments includes: S241, setting a plurality of divergent frames corresponding to the water conservancy estimation model, the number of the divergent frames being the same as the number of target learning models in the water conservancy estimation model, wherein the divergent frames include a plurality of guiding replication points; S242, setting a corresponding number of virtual spaces corresponding to the number of divergent frames; S243, the water conservancy calculation model is decrementally copied through the divergence frame to obtain multiple twin water conservancy calculation models, and the water conservancy calculation model is bound to the digital twin model; S244, storing multiple twin water conservancy calculation models one-to-one in corresponding virtual spaces through divergence boxes to obtain multiple calculation environments; As described in the above steps S241-S244, multiple divergent frames are set, wherein the divergent frames are multiple copy windows, and multiple twin water conservancy calculation models are obtained by copying the divergent frames and the water conservancy calculation model. The number of twin water conservancy calculation models is the same as the number of divergent frames, and the number of divergent frames is the same as the number of target learning models in the water conservancy calculation model. Then, a corresponding number of virtual spaces are set according to the number of divergent frames, and the water conservancy calculation model is decrementally copied through the divergent frames to obtain multiple twin water conservancy calculation models. The water conservancy calculation model is bound to the digital twin model, and the water quality data in the data twin model is obtained through the water conservancy calculation model. The water quality information obtained can be obtained through the water conservancy calculation model, and multiple twin water conservancy calculation models are stored one-to-one in the corresponding virtual space through the divergent frames to obtain multiple calculation environments. The finally formed connection relationship is to use the water conservancy calculation model as the starting point to communicate with the divergent frame, and at the same time, the divergent frame is connected to the corresponding virtual space, and the virtual space storing the twin water conservancy calculation model is used as the calculation environment.

[0022] In one embodiment, the step S243 of obtaining a plurality of twin water conservancy estimation models by decrementally replicating the water conservancy estimation model through a divergent frame includes: S2431, arranging multiple divergent frames, enabling multiple guide copy points in the divergent frame at the first position in the arrangement order, the number of the multiple guide copy points being the same as the number of target learning models in the water conservancy estimation model, and decreasing the number of enabled guide copy points by one for the divergent frames after the arrangement order; S2432, by using the guide replication points enabled in the multiple divergent boxes, sequentially and descendingly replicate the target learning model in the water conservancy estimation model to obtain multiple twin water conservancy estimation models; S2433, establishing perception channels between the water conservancy calculation model and multiple twin water conservancy calculation models based on the guided replication points; In one embodiment, step S2433 of establishing perception channels between the hydraulic estimation model and multiple twin hydraulic estimation models based on the guide replication points includes: S24331, the guiding replication point includes a sensing space, a plurality of sensing points in the sensing space, and water quality information configured at the sensing points; S24332. Establish a transmission channel between the water conservancy calculation model and the first target learning model upstream of the multiple twin water conservancy calculation models based on the guided replication point, connect the sensing point with the target learning model in the water conservancy calculation model that is the same as the first target learning model upstream of the twin water conservancy calculation model, and set the sensing attraction point and the water quality information configured by the sensing attraction point in the target learning model that is the same as the first target learning model upstream of the twin water conservancy calculation model. The sensing point will move toward the sensing attraction point with the same water quality information, and the sensing space will communicate with the first target learning model upstream of the multiple twin water conservancy calculation models to obtain a perception channel between the water conservancy calculation model and the replicated target learning models in the multiple twin water conservancy calculation models. As described in the above steps S2431-S2433, the divergent frames are arranged, and the guide copy points in the divergent frames are enabled in descending order. The guide copy points enabled in the first divergent frame are the same as the number of target learning models in the water conservancy calculation model. Therefore, the copying starts from copying the target learning model in the entire water conservancy calculation model in a descending order, and the target learning model corresponding to the upstream monitoring point is copied in descending order through the guide copy points enabled in multiple divergent frames to obtain multiple twin water conservancy calculation models. The copying of the target learning model in the water conservancy calculation model by the guide copy points is a descending copy of the target learning model corresponding to the upstream monitoring point. The guide copy points include The functions of copying and guiding are as follows: copying is to copy the target learning model in the water conservancy calculation model; guiding is to obtain data through the sensing space, multiple sensing points in the sensing space, and water quality information configured at the sensing points. Based on the guided copy points, a transmission channel is established between the water conservancy calculation model and the first upstream target learning model in multiple twin water conservancy calculation models; the sensing points are connected to the target learning model in the water conservancy calculation model that is the same as the first upstream target learning model in the twin water conservancy calculation model; and the sensing attraction points and the water quality information configured at the sensing attraction points are set in the target learning model that is the same as the first upstream target learning model in the twin water conservancy calculation model. The sensing points will move toward the sensing points with the same water quality information. The attraction point moves, and the sensing space communicates and connects with the first target learning model upstream of multiple twin water conservancy calculation models, so as to obtain a perception channel between the water conservancy calculation model and the target learning models copied in multiple twin water conservancy calculation models. For example, the water conservancy calculation model obtains the water quality information collected by the monitoring points in the bound digital twin model, and sets the sensing attraction point and the water quality information configured by the sensing attraction point in the target learning model in the water conservancy calculation model. Moreover, since the sensing point is connected to the target learning model in the water conservancy calculation model that is the same as the first target learning model upstream of the twin water conservancy calculation model, when the water conservancy calculation model obtains the water quality information collected by the monitoring points in the bound digital twin model, the corresponding water quality can be deduced. The sensing attraction points of the information are connected, and since the sensing points are connected to the first target learning model upstream in the water conservancy calculation model, the sensing points in the transmission channel can be attracted and connected by the corresponding sensing attraction points and leave the sensing space. When the water quality information in the water conservancy calculation model changes, the previous sensing attraction points will be withdrawn, and the corresponding sensing points can return to the sensing space. The sensing space is communicated and connected with the first target learning model upstream in multiple twin water conservancy calculation models. The first target learning model upstream in multiple twin water conservancy calculation models is connected to all sensing points in the sensing space. Therefore, the twin water conservancy calculation model can perceive the sensing points missing in the sensing space (the twin water conservancy calculation model is disconnected from the sensing points in the sensing space,The sensing point must exist in the sensing space to be connected to the twin water conservancy calculation model. The water quality information of the missing sensing points is configured to obtain water quality information in the water conservancy calculation model. It can be quickly acquired by the twin water conservancy calculation model, and the efficiency of the transmission of water quality information collected by the monitoring point can be improved. When the water quality information changes in a short period of time, it can be quickly acquired and changed by the twin water conservancy calculation model to ensure the timeliness and adaptability of the twin water conservancy calculation model.

[0023] In one embodiment, the step S3 of collecting water conservancy data in real time based on water conservancy object information and updating the digital twin model in real time, and calculating the water quality data of the water conservancy object according to the water conservancy calculation model to obtain target water quality development prediction information includes: S31. Collect water quality information in real time according to the monitoring points and update the water quality information into the digital twin model; S32, obtaining water quality information in the corresponding digital twin model through the hydraulic calculation model, the hydraulic calculation model derives the corresponding sensing attraction point according to the water quality information of the monitoring point, and attracts the sensing point corresponding to the water quality information through the transmission channel between the hydraulic calculation model and the first target learning model upstream of the multiple twin hydraulic calculation models; S33. When the sensing point is out of the sensing space, the first target learning model upstream of the twin water conservancy calculation model will calculate the water quality information corresponding to the sensing point that is missing in the sensing space, and will gather the twin water conservancy calculation models with overlapping water quality development prediction information, and will use the water quality development prediction information corresponding to the largest number of twin water conservancy calculation models in the collection as the target water quality development prediction information.

[0024] As described in the above steps S31-S33, water quality information is collected in real time according to the monitoring points, and the water quality information is updated to the digital twin model. The water quality information in the corresponding digital twin model is synchronously obtained through the hydraulic calculation model. The hydraulic calculation model deduces the sensing attraction points of the corresponding water quality information according to the water quality information of the monitoring points, and attracts the sensing points of the corresponding water quality information through the transmission channel between the hydraulic calculation model and the first target learning model upstream of the multiple twin hydraulic calculation models. When the sensing point is out of the sensing space, it cannot be connected to the first target learning model upstream of the twin hydraulic calculation model, and can be perceived by the first target learning model upstream of the twin hydraulic calculation model. The water quality information corresponding to the sensing points that are missing in the sensing space is put into the corresponding twin water conservancy calculation model for calculation. Since there is only water quality information of the first upstream monitoring point, it can be calculated downward through the twin water conservancy calculation model. The twin water conservancy calculation models with overlapping water quality development prediction information are grouped together, which means that the twin water conservancy calculation model with repeated calculation results is more accurate. Then, the corresponding water quality development prediction information of the largest number of twin water conservancy calculation models is taken as the target water quality development prediction information. Multiple twin water conservancy calculation models are independent of each other, which can improve the long-term calculation information of water quality information and facilitate the staff to carry out subsequent water quality maintenance.

[0025] Example 2, please refer to Figure 2 As shown, the digital twin water conservancy monitoring system based on artificial intelligence described in this embodiment includes: A construction module is used to determine water conservancy object information, wherein the water conservancy object information includes the water conservancy monitoring scope and the monitoring parameter type, and to construct a digital twin model according to the water conservancy object information; The setting module is connected with the construction module and is used to bind the water conservancy calculation model to the corresponding digital twin model, and obtain multiple calculation environments based on the water conservancy calculation model; The calculation module is connected to the setting module, and is used to collect water conservancy data in real time based on the water conservancy object information and update the digital twin model in real time. The water quality data of the water conservancy object is calculated according to the water conservancy calculation model to obtain the target water quality development prediction information.

[0026] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A digital twin water conservancy monitoring method based on artificial intelligence, characterized in that: The following steps are involved: Determine water conservancy object information, where the water conservancy object information includes water conservancy monitoring scope and monitoring parameter type, and build a digital twin model based on the water conservancy object information; The corresponding digital twin model is bound to the water conservancy calculation model, and multiple calculation environments are obtained based on the divergence of the water conservancy calculation model; Based on the information of water conservancy objects, water conservancy data is collected in real time and the digital twin model is updated in real time. The water quality data of the water conservancy objects is calculated according to the water conservancy calculation model to obtain the target water quality development prediction information.

2. The digital twin water conservancy monitoring method based on artificial intelligence according to claim 1 is characterized in that: The step of constructing a digital twin model according to water conservancy object information includes: Define the water conservancy monitoring scope, determine the types of monitoring parameters that need to be collected within the water conservancy monitoring scope, and use the water conservancy monitoring scope and monitoring parameter types as water conservancy object information; Set up multiple monitoring points within the water conservancy monitoring range corresponding to the monitoring parameter types; A water conservancy model is constructed according to the water conservancy monitoring scope, corresponding monitoring points are marked in the water conservancy model to obtain marking points, and corresponding monitoring scopes are divided according to corresponding marking points on the water conservancy model; The corresponding cloud server is configured for the corresponding marking point, and the cloud server is connected to the corresponding monitoring point for communication to obtain a digital twin model.

3. The digital twin water conservancy monitoring method based on artificial intelligence according to claim 1 is characterized in that: The corresponding digital twin model is bound to a water conservancy calculation model, and the step of diverging and obtaining multiple calculation environments based on the water conservancy calculation model includes: Obtain the historical water quality information of each monitoring point within the water conservancy monitoring range, and connect the monitoring points according to the water flow direction within the water conservancy monitoring range to obtain a monitoring point chain; Corresponding learning models are configured between adjacent monitoring points in the monitoring point chain, and the learning models corresponding to the monitoring points are trained through the historical water quality information of the adjacent monitoring points to obtain the target learning model; Based on the monitoring point chain, the target learning models corresponding to the adjacent monitoring points are arranged and connected to obtain the water conservancy estimation model; The water conservancy calculation model is diverged to obtain multiple calculation environments.

4. The artificial intelligence-based digital twin water conservancy monitoring method according to claim 3 is characterized in that: The step of diverging the water conservancy estimation model to obtain multiple estimation environments includes: A plurality of divergence boxes are set corresponding to the water conservancy estimation model, and the number of the divergence boxes is the same as the number of the target learning models in the water conservancy estimation model, wherein the divergence box includes a plurality of guide replication points; Set a corresponding number of virtual spaces according to the number of divergent frames; The water conservancy calculation model is copied incrementally through the divergence frame to obtain multiple twin water conservancy calculation models, and the binding between the water conservancy calculation model and the digital twin model is retained; Multiple twin water conservancy calculation models are stored one-to-one in the corresponding virtual space through a divergent frame to obtain multiple calculation environments.

5. The artificial intelligence-based digital twin water conservancy monitoring method according to claim 4 is characterized in that: The step of obtaining a plurality of twin water conservancy estimation models by decrementally replicating the water conservancy estimation model through a divergent frame includes: Arrange multiple divergent frames, enable multiple guide replication points in the first divergent frame in the arrangement order, the number of the multiple guide replication points is the same as the number of target learning models in the water conservancy estimation model, and reduce the number of enabled guide replication points by one for the divergent frames after the arrangement order; The target learning model in the water conservancy estimation model is sequentially and descendingly replicated through the guided replication points enabled in the multiple divergent boxes to obtain multiple twin water conservancy estimation models; Based on the guided replication points, perception channels are established between the water conservancy calculation model and multiple twin water conservancy calculation models.

6. The artificial intelligence-based digital twin water conservancy monitoring method according to claim 5 is characterized in that: The steps of establishing perception channels between the water conservancy estimation model and multiple twin water conservancy estimation models based on the guided replication points include: The guided replication point includes a sensing space, a plurality of sensing points in the sensing space, and water quality information configured at the sensing points; Based on the guided replication point, a transmission channel is established between the hydraulic calculation model and the first upstream target learning model in multiple twin hydraulic calculation models, the sensing point is connected to the target learning model in the hydraulic calculation model that is the same as the first upstream target learning model in the twin hydraulic calculation model, and the sensing attraction point and the water quality information configured by the sensing attraction point are set in the target learning model that is the same as the first upstream target learning model in the twin hydraulic calculation model. The sensing point will move toward the sensing attraction point with the same water quality information, and the sensing space is communicated and connected with the first upstream target learning model in multiple twin hydraulic calculation models to obtain a perception channel between the hydraulic calculation model and the replicated target learning models in multiple twin hydraulic calculation models.

7. The artificial intelligence-based digital twin water conservancy monitoring method according to claim 6 is characterized in that: The step of collecting water conservancy data in real time based on water conservancy object information and updating the digital twin model in real time, and calculating the water quality data of the water conservancy object according to the water conservancy calculation model to obtain target water quality development prediction information includes: Collect water quality information in real time based on monitoring points and update the water quality information to the digital twin model; The water quality information in the corresponding digital twin model is obtained through the hydraulic calculation model. The hydraulic calculation model derives the corresponding sensing attraction point according to the water quality information of the monitoring point, and attracts the sensing point corresponding to the water quality information through the transmission channel between the hydraulic calculation model and the first target learning model upstream of the multiple twin hydraulic calculation models; When the sensing point is out of the sensing space, the first target learning model upstream of the twin water conservancy calculation model will calculate the water quality information corresponding to the sensing points that are missing in the sensing space, and will collect the twin water conservancy calculation models with overlapping water quality development prediction information, and will use the corresponding water quality development prediction information of the twin water conservancy calculation models with the largest number of collected twin water conservancy calculation models as the target water quality development prediction information.

8. An artificial intelligence-based digital twin water conservancy monitoring system, used to implement the artificial intelligence-based digital twin water conservancy monitoring method according to any one of claims 1 to 7, characterized in that: include: A construction module is used to determine water conservancy object information, wherein the water conservancy object information includes the water conservancy monitoring scope and the monitoring parameter type, and to construct a digital twin model according to the water conservancy object information; The setting module is connected with the construction module and is used to bind the water conservancy calculation model to the corresponding digital twin model, and obtain multiple calculation environments based on the water conservancy calculation model; The calculation module is connected to the setting module, and is used to collect water conservancy data in real time based on the water conservancy object information and update the digital twin model in real time. The water quality data of the water conservancy object is calculated according to the water conservancy calculation model to obtain the target water quality development prediction information.

Images