Underground water level monitoring method and system based on artificial intelligence

Abstract

The embodiment of the invention provides a groundwater level monitoring method and system based on artificial intelligence, and the method comprises the steps: constructing a first groundwater level prediction network which comprises a focusing node and a transmission node, and generating a plurality of second groundwater level prediction networks based on the incidence relation of different nodes; and the diversity and accuracy of underground water level prediction are obviously improved. And linkage parameter learning is performed on the plurality of second groundwater level prediction networks by using the groundwater dynamic monitoring data of the basic sample, so that the network parameters are effectively optimized, and the prediction precision is further improved. Meanwhile, more abundant and comprehensive target sample groundwater dynamic monitoring data are generated by performing water level prediction learning sample extension on the basic sample data, and data support is provided for training of a target groundwater level prediction network. The finally generated target groundwater level prediction network is excellent in performance in practical application, and an accurate and reliable decision basis can be provided for groundwater management and water resource protection.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and more specifically, to a groundwater level monitoring method and system based on artificial intelligence. Background Art

[0002] As an important water resource, accurate monitoring of groundwater levels is of great significance for water resource management, environmental protection, and disaster warning. However, although traditional groundwater level monitoring methods, such as artificial observation wells and automatic recorders, can provide certain monitoring data, they have problems such as uneven distribution of monitoring points, untimely data updates, and high monitoring costs, making it difficult to meet the needs of groundwater level monitoring over a large range and over a long time scale. Summary of the invention

[0003] In view of this, the purpose of this application is to provide a groundwater level monitoring method and system based on artificial intelligence.

[0004] In conjunction with the first aspect of the present application, a groundwater level monitoring method based on artificial intelligence is provided, which is applied to a groundwater level monitoring system based on artificial intelligence, and the method comprises: Acquire a first groundwater level prediction network, wherein the first groundwater level prediction network includes a plurality of deep learning nodes, and the deep learning nodes include a focusing node and a transmission node; Generate multiple second groundwater level prediction networks based on different node association relationships between the focus node and the transmission node in each of the deep learning nodes; Based on the basic sample groundwater dynamic monitoring data, linkage parameter learning is performed on the plurality of second groundwater level prediction networks to generate a plurality of second groundwater level prediction networks that have completed parameter learning; Expanding the water level prediction learning samples of the basic sample groundwater dynamic monitoring data according to the plurality of second groundwater level prediction networks that have completed parameter learning, and generating target sample groundwater dynamic monitoring data; The first groundwater level prediction network is subjected to water level prediction learning based on the target sample groundwater dynamic monitoring data to generate a target groundwater level prediction network.

[0005] In a possible implementation of the first aspect, the generating of multiple second groundwater level prediction networks based on different node association relationships between the focus node and the transfer node in each of the deep learning nodes includes: Obtaining the scale of the deep learning nodes in the first groundwater level prediction network; Determining a plurality of network model architectures according to the scale of the deep learning node, each of the network model architectures corresponding to a node association relationship between a focusing node and a transmission node; A plurality of second groundwater level prediction networks are generated based on the plurality of network model architectures.

[0006] In a possible implementation of the first aspect, the method of performing linkage parameter learning on the plurality of second groundwater level prediction networks based on the basic sample groundwater dynamic monitoring data to generate a plurality of second groundwater level prediction networks that have completed parameter learning includes: Obtaining basic sample groundwater dynamic monitoring data, wherein the basic sample groundwater dynamic monitoring data includes a plurality of sample groundwater dynamic monitoring data and annotated estimated water level data corresponding to each sample groundwater dynamic monitoring data; Loading any target sample groundwater dynamic monitoring data from the multiple sample groundwater dynamic monitoring data into the multiple second groundwater level prediction networks respectively, and generating multiple predicted water level results corresponding to the target sample groundwater dynamic monitoring data; Optimizing the plurality of second groundwater level prediction networks based on errors between the plurality of predicted water level results and the annotated estimated water level data corresponding to the target sample groundwater dynamic monitoring data, wherein the plurality of second groundwater level prediction networks share neuron weight information; When the plurality of second groundwater level prediction networks do not meet the training termination condition, return to the step of loading any target sample groundwater dynamic monitoring data into the plurality of second groundwater level prediction networks respectively, and optimizing the neuron weight information of the plurality of second groundwater level prediction networks based on the error between the output plurality of predicted water level results and the corresponding labeled estimated water level data; When the plurality of second groundwater level prediction networks meet the training termination condition, a plurality of second groundwater level prediction networks that have completed parameter learning are generated.

[0007] In a possible implementation of the first aspect, the multiple second groundwater level prediction networks are optimized based on the errors between the multiple predicted water level results and the labeled estimated water level data corresponding to the target sample groundwater dynamic monitoring data, and the multiple second groundwater level prediction networks share neuron weight information, including: Respectively calculating the loss function value between each of the multiple predicted water level results and the labeled estimated water level data corresponding to the target sample groundwater dynamic monitoring data to generate multiple training cost results; Calculate the training direction parameters corresponding to each training cost result and generate multiple training direction parameters; The plurality of second groundwater level prediction networks are optimized based on the plurality of training direction parameters, and the plurality of second groundwater level prediction networks share neuron weight information.

[0008] In a possible implementation of the first aspect, the step of performing water level prediction learning sample expansion on the basic sample groundwater dynamic monitoring data based on the multiple second groundwater level prediction networks that have completed parameter learning to generate target sample groundwater dynamic monitoring data includes: Expanding the basic sample groundwater dynamic monitoring data based on the plurality of second groundwater level prediction networks that have completed parameter learning to generate an expanded sample groundwater dynamic monitoring data sequence; A plurality of target extended sample groundwater dynamic monitoring data are determined from the extended sample groundwater dynamic monitoring data sequence, and the plurality of target extended sample groundwater dynamic monitoring data are loaded into the basic sample groundwater dynamic monitoring data to generate target sample groundwater dynamic monitoring data.

[0009] In a possible implementation of the first aspect, the multiple second groundwater level prediction networks that have completed parameter learning expand the basic sample groundwater dynamic monitoring data to generate an extended sample groundwater dynamic monitoring data sequence, including: Processing each sample groundwater dynamic monitoring data in the basic sample groundwater dynamic monitoring data based on each second groundwater level prediction network that has completed parameter learning, and generating a confidence sequence predicted by each second groundwater level prediction network that has completed parameter learning; The extended sample groundwater dynamic monitoring data sequence is determined based on the confidence sequence output by the second groundwater level prediction network that has completed parameter learning.

[0010] In a possible implementation of the first aspect, the step of determining an extended sample groundwater dynamic monitoring data sequence based on the confidence sequence output by each second groundwater level prediction network that has completed parameter learning includes: Determine the maximum confidence in each confidence sequence, and generate the maximum confidence of each second groundwater level prediction network prediction that completes parameter learning; Determine the extended sample groundwater dynamic monitoring data corresponding to each maximum confidence level, and generate multiple extended sample groundwater dynamic monitoring data; An extended sample groundwater dynamic monitoring data sequence is determined based on the multiple extended sample groundwater dynamic monitoring data.

[0011] In a possible implementation of the first aspect, the method includes: Based on the target groundwater level prediction network, any input target groundwater dynamic monitoring data is predicted to generate a water level prediction result of the target groundwater dynamic monitoring data.

[0012] In combination with the second aspect of the present application, a groundwater level monitoring system based on artificial intelligence is provided, wherein the groundwater level monitoring system based on artificial intelligence includes a machine-readable storage medium and a processor, wherein the machine-readable storage medium stores machine-executable instructions, and when the processor executes the machine-executable instructions, the groundwater level monitoring system based on artificial intelligence implements the aforementioned groundwater level monitoring method based on artificial intelligence.

[0013] In conjunction with the third aspect of the present application, a computer-readable storage medium is provided, in which computer-executable instructions are stored. When the computer-executable instructions are executed, the aforementioned artificial intelligence-based groundwater level monitoring method is implemented.

[0014] In combination with any of the above aspects, the embodiment of the present application significantly improves the diversity and accuracy of groundwater level prediction by constructing a first groundwater level prediction network including focus nodes and transfer nodes, and generating multiple second groundwater level prediction networks based on the association relationships between different nodes. The basic sample groundwater dynamic monitoring data is used to perform linkage parameter learning on multiple second groundwater level prediction networks, which effectively optimizes the network parameters and further improves the prediction accuracy. At the same time, by expanding the water level prediction learning samples of the basic sample data, more abundant and comprehensive target sample groundwater dynamic monitoring data is generated, providing data support for the training of the target groundwater level prediction network. The target groundwater level prediction network finally generated performs well in practical applications and can provide accurate and reliable decision-making basis for groundwater management and water resource protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained by combining these drawings without paying creative work.

[0016] Figure 1 A schematic flow chart of a groundwater level monitoring method based on artificial intelligence provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, 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.

[0018] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0019] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0020] Figure 1 The flowchart of the groundwater level monitoring method based on artificial intelligence provided by the embodiment of the present application is shown. It should be understood that in other embodiments, the order of some steps of the groundwater level monitoring method based on artificial intelligence of this embodiment can be shared with each other based on actual needs, or some steps can be omitted or maintained. The groundwater level monitoring method based on artificial intelligence includes: Step S110, obtaining a first groundwater level prediction network, wherein the first groundwater level prediction network includes a plurality of deep learning nodes, and the deep learning nodes include a focusing node and a transmission node.

[0021] In this embodiment, when the server performs the groundwater level prediction task, it first needs to obtain the first groundwater level prediction network. This first groundwater level pre-network is the basic architecture of the entire prediction work. For example, the server obtains this first groundwater level pre-network from a pre-built model library. The deep learning node in this first groundwater level pre-network is the basic unit for building the network, in which the focusing node and the transmission node have their own functions. The focusing node may be responsible for centralized processing of specific data features, such as focusing on and processing seasonal variation features in groundwater level data. The transmission node is responsible for transmitting the data processed by the focusing node or the original data to other parts of the network to ensure the flow of data and information transmission in the entire network. Just like in a complex logistics network, the focusing node is like a site for special processing of special goods (specific data features), and the transmission node is like a transportation route connecting various sites (nodes) to ensure that the goods (data) can reach different parts of the network. The first groundwater level prediction network obtained by the server may have been preliminarily constructed and designed before, and its structure and node settings are set to adapt to the specific task of groundwater level prediction.

[0022] Step S120: generating a plurality of second groundwater level prediction networks based on different node association relationships between the focus node and the transmission node in each of the deep learning nodes.

[0023] After obtaining the first groundwater level prediction network, multiple second groundwater level prediction networks are constructed based on the different node associations between the focus nodes and the transmission nodes of the deep learning nodes. First, the server obtains the scale of the deep learning nodes in the first groundwater level prediction network. Assume that the scale of the deep learning nodes in this first network is 100, of which there are 30 focus nodes and 70 transmission nodes. The server determines multiple network model architectures based on this scale. For example, if the focus node and the transmission node are a one-to-one connection relationship, a network model architecture is determined; if it is a relationship in which a focus node is connected to multiple transmission nodes, another network model architecture is determined.

[0024] Based on these determined different network model architectures, the server generates multiple second groundwater level prediction networks. For example, when three different node associations are determined, the server will generate three different second groundwater level prediction networks. Taking one of the networks as an example, in this network, the association between the focus node and the transmission node is set so that the focus node first processes the long-term trend data in the groundwater level data, and then the transmission node passes the processed data to the next group of nodes for further analysis, such as analyzing the impact of seasonal fluctuations on the water level. This method of generating multiple networks based on different node associations can predict groundwater levels from multiple angles and structures, increasing the potential for prediction diversity and accuracy.

[0025] Step S130 , performing linkage parameter learning on the plurality of second groundwater level prediction networks based on the basic sample groundwater dynamic monitoring data, and generating a plurality of second groundwater level prediction networks that have completed parameter learning.

[0026] The server obtains basic sample groundwater dynamic monitoring data, which is pre-collected and sorted for training the second groundwater level prediction network. Assume that the basic sample groundwater dynamic monitoring data contains 1,000 sets of groundwater dynamic monitoring data from different regions and different time periods, and each set of data has corresponding annotated estimated water level data.

[0027] The server loads any target sample groundwater dynamic monitoring data from the 1,000 sets of sample groundwater dynamic monitoring data into multiple second groundwater level prediction networks. For example, a set of data from a certain period of time in a specific area is selected and loaded into each second groundwater level prediction network. Each network processes the data according to its own structure and algorithm to generate multiple predicted water level results corresponding to the target sample groundwater dynamic monitoring data.

[0028] Then, the server optimizes multiple second groundwater level prediction networks based on the errors between these predicted water level results and the labeled estimated water level data corresponding to the target sample groundwater dynamic monitoring data. Since these second groundwater level prediction networks share neuron weight information, the optimization is linked. Specifically, the server calculates the loss function value between each predicted water level result in multiple predicted water level results and the labeled estimated water level data corresponding to the target sample groundwater dynamic monitoring data, and generates multiple training cost results. For example, the gap between each predicted result and the true labeled value is calculated through a loss function such as the mean square error (MSE).

[0029] Then, the training direction parameters corresponding to each training cost result are calculated to generate multiple training direction parameters. These training direction parameters indicate the direction and amplitude in which the network parameters should be adjusted. Based on these training direction parameters, multiple second groundwater level prediction networks are optimized. If in this process, multiple second groundwater level prediction networks do not meet the training termination conditions, for example, the preset training rounds have not been completed or the prediction error has not reached the predetermined threshold, then the server will return to execute the steps of loading any target sample groundwater dynamic monitoring data into multiple second groundwater level prediction networks respectively, and optimizing the neuron weight information of multiple second groundwater level prediction networks based on the error between the output multiple predicted water level results and the corresponding labeled estimated water level data.

[0030] When multiple second groundwater level prediction networks meet the training termination conditions, for example, after 1,000 rounds of training, the prediction error has been reduced to an acceptable range, the server generates multiple second groundwater level prediction networks that have completed parameter learning. These networks have adjusted their own parameters by learning basic sample groundwater dynamic monitoring data, and have the ability to make more accurate predictions of groundwater levels.

[0031] Step S140, performing water level prediction learning sample expansion on the basic sample groundwater dynamic monitoring data according to the plurality of second groundwater level prediction networks that have completed parameter learning, to generate target sample groundwater dynamic monitoring data.

[0032] The server expands the basic sample groundwater dynamic monitoring data based on multiple second groundwater level prediction networks that have completed parameter learning. First, the server processes each sample groundwater dynamic monitoring data in the basic sample groundwater dynamic monitoring data based on each second groundwater level prediction network that has completed parameter learning, and generates a confidence sequence of predictions of each second groundwater level prediction network that has completed parameter learning. For example, for a certain group of groundwater dynamic monitoring data in the basic sample, each second groundwater level prediction network that has completed parameter learning will give a prediction result and the confidence of this result. Assuming that for 1,000 groups of basic sample data, each network will generate 1,000 confidence values, and these values ​​constitute a confidence sequence.

[0033] Then, the server determines the extended sample groundwater dynamic monitoring data sequence based on the confidence sequence output by each second groundwater level prediction network that has completed parameter learning. Specifically, the server determines the maximum confidence in each confidence sequence and generates the maximum confidence predicted by each second groundwater level prediction network that has completed parameter learning. For example, in the confidence sequence of a certain network, the prediction confidence of the 500th group of data is the highest, then this highest confidence is the maximum confidence of this network for this group of data.

[0034] Then, the extended sample groundwater dynamic monitoring data corresponding to each maximum confidence level is determined, and multiple extended sample groundwater dynamic monitoring data are generated. For example, the set of data corresponding to the maximum confidence level is found and used as the extended sample groundwater dynamic monitoring data. Finally, the extended sample groundwater dynamic monitoring data sequence is determined based on these multiple extended sample groundwater dynamic monitoring data.

[0035] Multiple target extended sample groundwater dynamic monitoring data are determined from this extended sample groundwater dynamic monitoring data sequence. Assuming that 100 sets of data are selected from the extended sample data sequence, these 100 sets of target extended sample groundwater dynamic monitoring data are loaded into the basic sample groundwater dynamic monitoring data to generate target sample groundwater dynamic monitoring data. These target sample groundwater dynamic monitoring data contain more valuable information and can provide richer data support for the subsequent learning of the first groundwater level prediction network.

[0036] Step S150, performing water level prediction learning on the first groundwater level prediction network according to the target sample groundwater dynamic monitoring data to generate a target groundwater level prediction network.

[0037] The server uses the generated target sample groundwater dynamic monitoring data to learn the water level prediction of the first groundwater level prediction network. Each set of data in the target sample groundwater dynamic monitoring data contains rich groundwater dynamic information and corresponding water level information. The first groundwater level prediction network adjusts its own parameters by learning these data to improve its ability to predict groundwater levels.

[0038] During the learning process, the focus nodes and transfer nodes in the network work together. The focus nodes extract and analyze the key features in the target sample data, such as focusing on the relationship between groundwater levels and rainfall, surrounding land use types and other factors. The transfer nodes pass these processed data to other parts of the network, allowing the entire network to adjust parameters such as the connection weights between neurons based on these data.

[0039] With the continuous learning of the target sample groundwater dynamic monitoring data, the prediction ability of the first groundwater level prediction network gradually improves. When the learning process reaches the predetermined conditions, such as the prediction error is reduced to a very low level or the learning rounds reach a certain number, the server uses the learned first groundwater level prediction network as the target groundwater level prediction network. This target groundwater level prediction network has undergone multiple rounds of learning and optimization and has the ability to accurately predict the groundwater level. It can be used to predict any input target groundwater dynamic monitoring data and generate the water level prediction results of the target groundwater dynamic monitoring data. For example, when receiving new groundwater dynamic monitoring data of a certain area, this target groundwater level prediction network can accurately predict the groundwater level of the area based on the previously learned knowledge and rules.

[0040] Based on the above steps, the embodiment of the present application significantly improves the diversity and accuracy of groundwater level prediction by constructing a first groundwater level prediction network including focus nodes and transfer nodes, and generating multiple second groundwater level prediction networks based on the association relationships between different nodes. The basic sample groundwater dynamic monitoring data is used to perform linkage parameter learning on multiple second groundwater level prediction networks, which effectively optimizes the network parameters and further improves the prediction accuracy. At the same time, by expanding the water level prediction learning samples of the basic sample data, more abundant and comprehensive target sample groundwater dynamic monitoring data is generated, providing data support for the training of the target groundwater level prediction network. The target groundwater level prediction network finally generated performs well in practical applications and can provide accurate and reliable decision-making basis for groundwater management and water resource protection.

[0041] In a possible implementation manner, the generating of multiple second groundwater level prediction networks based on different node association relationships between the focus node and the transfer node in each of the deep learning nodes includes: Obtain the scale of the deep learning nodes in the first groundwater level prediction network.

[0042] Multiple network model architectures are determined according to the scale of the deep learning nodes, and each of the network model architectures corresponds to a node association relationship between a focusing node and a transmission node.

[0043] A plurality of second groundwater level prediction networks are generated based on the plurality of network model architectures.

[0044] In a possible implementation, the linking parameter learning of the plurality of second groundwater level prediction networks based on the basic sample groundwater dynamic monitoring data to generate a plurality of second groundwater level prediction networks that have completed parameter learning includes: Basic sample groundwater dynamic monitoring data are obtained, wherein the basic sample groundwater dynamic monitoring data includes a plurality of sample groundwater dynamic monitoring data and annotated estimated water level data corresponding to each sample groundwater dynamic monitoring data.

[0045] Any target sample groundwater dynamic monitoring data among the multiple sample groundwater dynamic monitoring data are loaded into the multiple second groundwater level prediction networks respectively to generate multiple predicted water level results corresponding to the target sample groundwater dynamic monitoring data.

[0046] The multiple second groundwater level prediction networks are optimized based on the errors between the multiple predicted water level results and the labeled estimated water level data corresponding to the target sample groundwater dynamic monitoring data, and the multiple second groundwater level prediction networks share neuron weight information.

[0047] When the multiple second groundwater level prediction networks do not meet the training termination conditions, return to the step of loading any target sample groundwater dynamic monitoring data into the multiple second groundwater level prediction networks respectively, and optimizing the neuron weight information of the multiple second groundwater level prediction networks based on the errors between the output multiple predicted water level results and the corresponding labeled estimated water level data.

[0048] When the plurality of second groundwater level prediction networks meet the training termination condition, a plurality of second groundwater level prediction networks that have completed parameter learning are generated.

[0049] In a possible implementation, the optimization of the plurality of second groundwater level prediction networks based on the errors between the plurality of predicted water level results and the annotated estimated water level data corresponding to the target sample groundwater dynamic monitoring data, wherein the plurality of second groundwater level prediction networks share neuron weight information, includes: The loss function value between each predicted water level result in the multiple predicted water level results and the labeled estimated water level data corresponding to the target sample groundwater dynamic monitoring data is calculated respectively to generate multiple training cost results.

[0050] The training direction parameters corresponding to each training cost result are calculated to generate multiple training direction parameters.

[0051] The plurality of second groundwater level prediction networks are optimized based on the plurality of training direction parameters, and the plurality of second groundwater level prediction networks share neuron weight information.

[0052] In a possible implementation, the water level prediction learning sample expansion of the basic sample groundwater dynamic monitoring data based on the multiple second groundwater level prediction networks that have completed parameter learning to generate target sample groundwater dynamic monitoring data includes: The basic sample groundwater dynamic monitoring data is expanded based on the multiple second groundwater level prediction networks that have completed parameter learning to generate an extended sample groundwater dynamic monitoring data sequence.

[0053] A plurality of target extended sample groundwater dynamic monitoring data are determined from the extended sample groundwater dynamic monitoring data sequence, and the plurality of target extended sample groundwater dynamic monitoring data are loaded into the basic sample groundwater dynamic monitoring data to generate target sample groundwater dynamic monitoring data.

[0054] In a possible implementation, the multiple second groundwater level prediction networks based on the completed parameter learning expand the basic sample groundwater dynamic monitoring data to generate an extended sample groundwater dynamic monitoring data sequence, including: Based on each second groundwater level prediction network that has completed parameter learning, each sample groundwater dynamic monitoring data in the basic sample groundwater dynamic monitoring data is processed to generate a confidence sequence predicted by each second groundwater level prediction network that has completed parameter learning.

[0055] The extended sample groundwater dynamic monitoring data sequence is determined based on the confidence sequence output by the second groundwater level prediction network that has completed parameter learning.

[0056] In a possible implementation, the step of determining an extended sample groundwater dynamic monitoring data sequence based on the confidence sequence output by each second groundwater level prediction network that has completed parameter learning includes: The maximum confidence in each confidence sequence is determined, and the maximum confidence of each second groundwater level prediction network prediction that has completed parameter learning is generated.

[0057] The extended sample groundwater dynamic monitoring data corresponding to each maximum confidence level is determined, and a plurality of extended sample groundwater dynamic monitoring data are generated.

[0058] An extended sample groundwater dynamic monitoring data sequence is determined based on the multiple extended sample groundwater dynamic monitoring data.

[0059] In one possible implementation, the method includes: Based on the target groundwater level prediction network, any input target groundwater dynamic monitoring data is predicted to generate a water level prediction result of the target groundwater dynamic monitoring data.

[0060] In the above embodiments, the artificial intelligence-based groundwater level monitoring system for executing the above method embodiments has at least one processor, a control module (chip set) coupled to at least one of the (at least one) processors, a memory coupled to the control module, a non-volatile memory (NVM) / storage device coupled to the control module, at least one load / output device coupled to the control module, and a network interface coupled to the control module.

[0061] The processor may include at least one single-core or multi-core processor, and the processor may include any combination of general-purpose processors or special-purpose processors (such as graphics processors, application processors, baseband processors, etc.). For some alternative implementations, the groundwater level monitoring system based on artificial intelligence can be used as the gateway described in the embodiments of the present application. Based on electronic devices.

[0062] For some alternative embodiments, the artificial intelligence-based groundwater level monitoring system may include at least one computer-readable medium (e.g., a memory or NVM / storage device) having instructions and at least one processor integrated with the at least one computer-readable medium and configured to execute the instructions to implement a module to perform the actions described in the present disclosure.

[0063] For one embodiment, the control module may include any suitable interface controller to provide any suitable interface to at least one of the processor(s) and / or any suitable device or component in communication with the control module.

[0064] The control module may include a memory controller module to provide an interface to the memory. The memory controller module may be a hardware module, a software module, and / or a firmware module.

[0065] The memory may be used, for example, to load and store data and / or instructions for an artificial intelligence-based groundwater level monitoring system. For one embodiment, the memory may include any suitable volatile memory, such as a suitable DRAM.

[0066] For one embodiment, the control module may include at least one load-in / load-out controller to provide an interface to the NVM / storage device and the (at least one) load-in / load-out device.

[0067] For example, the NVM / storage device may be used to store data and / or instructions. The NVM / storage device may include any suitable non-volatile memory (e.g., flash memory) and / or may include any suitable (at least one) non-volatile storage device (e.g., at least one hard disk drive (HDD), at least one compact disk (CD) drive, and / or at least one digital versatile disk (DVD) drive).

[0068] The NVM / storage device may include storage resources that are physically part of the device on which the artificial intelligence-based groundwater level monitoring system is installed, or it may be accessible to the device without being part of the device. For example, the NVM / storage device may be accessed via (at least one) load / output device based on a network.

[0069] The (at least one) load-in / output device may provide an interface for the artificial intelligence-based groundwater level monitoring system to communicate with any other appropriate device, and the load-in / output device may include a communication component, a phonetic component, a sensor component, etc. The network interface may provide an interface for the artificial intelligence-based groundwater level monitoring system to communicate based on at least one network, and the artificial intelligence-based groundwater level monitoring system may wirelessly communicate with at least one component of a wireless network based on any of at least one wireless network priors and / or protocols, such as accessing a wireless network based on a communication prior.

[0070] For one embodiment, at least one of the (at least one) processor may be loaded together with the logic of at least one controller of the control module (e.g., a memory controller module). For one embodiment, at least one of the (at least one) processor may be loaded together with the logic of at least one controller of the control module to form a system level load. For one embodiment, at least one of the (at least one) processor may be fused on the same die with the logic of at least one controller of the control module. For one embodiment, at least one of the (at least one) processor may be fused on the same die with the logic of at least one controller of the control module to form a system on chip (SoC).

[0071] The embodiments of the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, based on the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

[0072] An embodiment of the present invention discloses a computer-readable storage medium storing a computer program for electronic data exchange, wherein the computer program enables a computer to execute the steps of the groundwater level monitoring method based on artificial intelligence described in the aforementioned embodiment.

[0073] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the artificial intelligence-based groundwater level monitoring method described in the aforementioned embodiment.

[0074] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed over multiple modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. Those of ordinary skill in the art may understand and implement the present invention without creative effort.

[0075] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other medium that can be used to be computer-readable with or store data.

[0076] Finally, it should be noted that what is disclosed above is only a preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A groundwater level monitoring method based on artificial intelligence, characterized in that: The method comprises: Acquire a first groundwater level prediction network, wherein the first groundwater level prediction network includes a plurality of deep learning nodes, and the deep learning nodes include a focusing node and a transmission node; Generate multiple second groundwater level prediction networks based on different node association relationships between the focus node and the transmission node in each of the deep learning nodes; Based on the basic sample groundwater dynamic monitoring data, linkage parameter learning is performed on the plurality of second groundwater level prediction networks to generate a plurality of second groundwater level prediction networks that have completed parameter learning; Expanding the water level prediction learning samples of the basic sample groundwater dynamic monitoring data according to the plurality of second groundwater level prediction networks that have completed parameter learning, and generating target sample groundwater dynamic monitoring data; The first groundwater level prediction network is subjected to water level prediction learning based on the target sample groundwater dynamic monitoring data to generate a target groundwater level prediction network.

2. The groundwater level monitoring method based on artificial intelligence according to claim 1 is characterized in that: The generating of a plurality of second groundwater level prediction networks based on different node association relationships between the focus node and the transmission node in each of the deep learning nodes comprises: Obtaining the scale of the deep learning nodes in the first groundwater level prediction network; Determining a plurality of network model architectures according to the scale of the deep learning node, each of the network model architectures corresponding to a node association relationship between a focusing node and a transmission node; A plurality of second groundwater level prediction networks are generated based on the plurality of network model architectures.

3. The method for monitoring groundwater level based on artificial intelligence according to claim 1 or 2, characterized in that: The method of performing linkage parameter learning on the plurality of second groundwater level prediction networks based on the basic sample groundwater dynamic monitoring data to generate a plurality of second groundwater level prediction networks that have completed parameter learning includes: Obtaining basic sample groundwater dynamic monitoring data, wherein the basic sample groundwater dynamic monitoring data includes a plurality of sample groundwater dynamic monitoring data and annotated estimated water level data corresponding to each sample groundwater dynamic monitoring data; Loading any target sample groundwater dynamic monitoring data from the multiple sample groundwater dynamic monitoring data into the multiple second groundwater level prediction networks respectively, and generating multiple predicted water level results corresponding to the target sample groundwater dynamic monitoring data; Optimizing the plurality of second groundwater level prediction networks based on errors between the plurality of predicted water level results and the annotated estimated water level data corresponding to the target sample groundwater dynamic monitoring data, wherein the plurality of second groundwater level prediction networks share neuron weight information; When the plurality of second groundwater level prediction networks do not meet the training termination condition, return to the step of loading any target sample groundwater dynamic monitoring data into the plurality of second groundwater level prediction networks respectively, and optimizing the neuron weight information of the plurality of second groundwater level prediction networks based on the error between the output plurality of predicted water level results and the corresponding annotated estimated water level data; When the plurality of second groundwater level prediction networks meet the training termination condition, a plurality of second groundwater level prediction networks that have completed parameter learning are generated.

4. The groundwater level monitoring method based on artificial intelligence according to claim 3 is characterized in that: The optimizing the plurality of second groundwater level prediction networks based on the errors between the plurality of predicted water level results and the annotated estimated water level data corresponding to the target sample groundwater dynamic monitoring data, wherein the plurality of second groundwater level prediction networks share neuron weight information, comprises: Respectively calculating the loss function value between each of the multiple predicted water level results and the labeled estimated water level data corresponding to the target sample groundwater dynamic monitoring data to generate multiple training cost results; Calculate the training direction parameters corresponding to each training cost result and generate multiple training direction parameters; The plurality of second groundwater level prediction networks are optimized based on the plurality of training direction parameters, and the plurality of second groundwater level prediction networks share neuron weight information.

5. The groundwater level monitoring method based on artificial intelligence according to claim 1 is characterized in that: The step of performing water level prediction learning sample expansion on the basic sample groundwater dynamic monitoring data based on the plurality of second groundwater level prediction networks that have completed parameter learning to generate target sample groundwater dynamic monitoring data includes: Expanding the basic sample groundwater dynamic monitoring data based on the plurality of second groundwater level prediction networks that have completed parameter learning to generate an expanded sample groundwater dynamic monitoring data sequence; A plurality of target extended sample groundwater dynamic monitoring data are determined from the extended sample groundwater dynamic monitoring data sequence, and the plurality of target extended sample groundwater dynamic monitoring data are loaded into the basic sample groundwater dynamic monitoring data to generate target sample groundwater dynamic monitoring data.

6. The groundwater level monitoring method based on artificial intelligence according to claim 5 is characterized in that: The method of expanding the basic sample groundwater dynamic monitoring data based on the plurality of second groundwater level prediction networks that have completed parameter learning to generate an extended sample groundwater dynamic monitoring data sequence includes: Processing each sample groundwater dynamic monitoring data in the basic sample groundwater dynamic monitoring data based on each second groundwater level prediction network that has completed parameter learning, and generating a confidence sequence predicted by each second groundwater level prediction network that has completed parameter learning; The extended sample groundwater dynamic monitoring data sequence is determined based on the confidence sequence output by the second groundwater level prediction network that has completed parameter learning.

7. The groundwater level monitoring method based on artificial intelligence according to claim 6 is characterized in that: The method of determining an extended sample groundwater dynamic monitoring data sequence based on the confidence sequence output by the second groundwater level prediction network that has completed parameter learning comprises: Determine the maximum confidence in each confidence sequence, and generate the maximum confidence of each second groundwater level prediction network prediction that completes parameter learning; Determine the extended sample groundwater dynamic monitoring data corresponding to each maximum confidence level, and generate multiple extended sample groundwater dynamic monitoring data; An extended sample groundwater dynamic monitoring data sequence is determined based on the multiple extended sample groundwater dynamic monitoring data.

8. The method for monitoring groundwater level based on artificial intelligence according to claim 1, characterized in that: The method comprises: Based on the target groundwater level prediction network, any input target groundwater dynamic monitoring data is predicted to generate a water level prediction result of the target groundwater dynamic monitoring data.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by a computer, the groundwater level monitoring method based on artificial intelligence described in any one of claims 1 to 8 is implemented.

10. A groundwater level monitoring system based on artificial intelligence, characterized in that: It includes a processor and a computer-readable storage medium, wherein the computer-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by a computer, the artificial intelligence-based groundwater level monitoring method described in any one of claims 1 to 8 is implemented.

Images