High-precision monitoring system and method for groundwater flow velocity and direction based on multi-parameter fusion technology

Abstract

The present application relates to the field of groundwater velocity and direction monitoring technology, and specifically discloses a high-precision monitoring system and method for groundwater velocity and direction based on multi-parameter fusion technology, which uses data analysis and processing technology based on deep learning to implicitly associate groundwater pressure and groundwater velocity in time series, and implicitly associate groundwater temperature values ​​in space, so as to automatically obtain the short-term prediction value of groundwater direction based on the fine-grained matching interaction representation of groundwater pressure time series correlation characteristics and groundwater velocity time series correlation characteristics and the interactive response fusion representation between the spatial implicit correlation characteristics of groundwater temperature values. In this way, by comprehensively analyzing multiple parameters such as groundwater pressure, velocity and temperature, the complexity of the groundwater system can be more comprehensively reflected, and the interaction between different parameters can be captured, so as to better understand the dynamic changes of the groundwater system and improve the accuracy of prediction.

Description

Technical Field

[0001] The present application relates to the technical field of groundwater velocity and direction monitoring, and more specifically, to a high-precision groundwater velocity and direction monitoring system and method based on multi-parameter fusion technology. Background Art

[0002] The groundwater system is an important part of the Earth's geological structure. Its flow direction and speed are of great significance for studying the structure, permeability and connectivity of underground rock formations. By monitoring and analyzing the flow of groundwater, we can gain a deeper understanding of the properties and distribution of underground rocks, thus providing valuable data for geological research.

[0003] However, traditional monitoring methods usually monitor and predict based on a single parameter (such as groundwater level, flow rate, etc.), ignoring the combined effects of other related parameters. Specifically, a single parameter cannot fully reflect the complexity of the groundwater system and cannot capture the interaction between different parameters. For example, monitoring the groundwater level alone cannot reflect the impact of pressure changes on flow rate, nor can it reflect the impact of temperature on permeability. In addition, the prediction model established based on a single parameter cannot reflect the interaction between multiple factors, resulting in inaccurate prediction results and difficulty in coping with the dynamic changes of the groundwater system, thereby limiting the effectiveness and practicality of monitoring.

[0004] Therefore, an optimized groundwater flow velocity and direction monitoring system is desired. Summary of the invention

[0005] The present application provides a high-precision monitoring system and method for groundwater flow velocity and direction based on multi-parameter fusion technology. By comprehensively analyzing multiple parameters such as groundwater pressure, flow velocity and temperature, it can more comprehensively reflect the complexity of the groundwater system and capture the interaction between different parameters, so as to better understand the dynamic changes of the groundwater system and improve the accuracy of prediction, thereby enhancing the practicality and intelligence level of the monitoring system.

[0006] In the first aspect, a high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology is provided, comprising:

[0007] A groundwater parameter acquisition module is used to obtain a time sequence of groundwater pressure, a time sequence of groundwater velocity, and a depth sequence of groundwater temperature values;

[0008] A groundwater parameter implicit association module is used to extract a groundwater pressure time series implicit association feature vector, a groundwater flow velocity time series implicit association feature vector and a groundwater temperature space implicit association feature vector from the groundwater pressure time series, the groundwater flow velocity time series and the groundwater temperature value depth sequence respectively;

[0009] A groundwater pressure-velocity fusion interaction module is used to perform fine-grained optimal fusion interaction on the groundwater pressure time series implicit correlation feature vector and the groundwater velocity time series implicit correlation feature vector to obtain a groundwater pressure-velocity time series collaborative fusion feature;

[0010] A groundwater spatiotemporal multi-parameter fusion module, used for performing local and global collaborative fusion interaction on the groundwater pressure-velocity time series collaborative fusion characteristics and the groundwater temperature spatial implicit correlation feature vector to obtain a groundwater spatiotemporal multi-parameter fusion representation, wherein the groundwater spatiotemporal multi-parameter fusion module includes: a groundwater temperature decoupling unit, used for performing feature decoupling on the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of groundwater temperature local spatial implicit correlation characteristics; a groundwater multi-parameter interactive fusion unit, used for performing feature interactive fusion on the groundwater pressure-velocity time series collaborative fusion characteristics and the sequence of groundwater temperature local spatial implicit correlation characteristics to obtain the groundwater spatiotemporal multi-parameter fusion representation;

[0011] The flow direction monitoring result generating module is used to obtain the flow direction monitoring result based on the groundwater spatiotemporal multi-parameter fusion representation.

[0012] In the second aspect, a high-precision monitoring method for groundwater flow velocity and direction based on multi-parameter fusion technology is provided, including:

[0013] Obtaining a time series of groundwater pressure, a time series of groundwater velocity, and a depth series of groundwater temperature values;

[0014] Extracting a groundwater pressure time series implicit correlation feature vector, a groundwater velocity time series implicit correlation feature vector and a groundwater temperature space implicit correlation feature vector from the groundwater pressure time series, the groundwater velocity time series and the groundwater temperature depth series respectively;

[0015] Performing fine-grained optimal fusion interaction on the implicit correlation feature vector of the groundwater pressure time series and the implicit correlation feature vector of the groundwater velocity time series to obtain a groundwater pressure-velocity time series collaborative fusion feature;

[0016] The groundwater pressure-velocity time series collaborative fusion feature and the groundwater temperature spatial implicit correlation feature vector are subjected to local and global collaborative fusion interaction to obtain a groundwater spatiotemporal multi-parameter fusion representation, wherein the groundwater pressure-velocity time series collaborative fusion feature and the groundwater temperature spatial implicit correlation feature vector are subjected to local and global collaborative fusion interaction, including: feature decoupling of the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of groundwater temperature local spatial implicit correlation features; feature interactive fusion of the groundwater pressure-velocity time series collaborative fusion feature and the sequence of groundwater temperature local spatial implicit correlation features to obtain the groundwater spatiotemporal multi-parameter fusion representation;

[0017] Based on the groundwater spatiotemporal multi-parameter fusion representation, the flow direction monitoring result is obtained.

[0018] The present application provides a high-precision monitoring system and method for groundwater flow velocity and direction based on multi-parameter fusion technology, which obtains the time sequence of groundwater pressure, the time sequence of groundwater flow velocity and the depth sequence of groundwater temperature value, and adopts data analysis and processing technology based on deep learning to implicitly associate the groundwater pressure and the groundwater flow velocity in time series, and implicitly associate the groundwater temperature value in space, so as to automatically obtain the short-term prediction value of groundwater flow direction according to the fine-grained matching interaction representation of the time series correlation characteristics of groundwater pressure and the time series correlation characteristics of groundwater flow velocity and the interactive response fusion representation between the spatial implicit correlation characteristics of groundwater temperature value. In this way, by comprehensively analyzing multiple parameters such as groundwater pressure, flow velocity and temperature, the complexity of the groundwater system can be more comprehensively reflected, and the interaction between different parameters can be captured, so as to better understand the dynamic changes of the groundwater system and improve the accuracy of prediction, thereby improving the practicality and intelligence level of the monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0020] Figure 1 This is a schematic block diagram of a high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology in an embodiment of the present application.

[0021] Figure 2 This is a schematic diagram of data flow in a high-precision monitoring system for groundwater flow velocity based on multi-parameter fusion technology in an embodiment of the present application.

[0022] Figure 3It is a schematic block diagram of a groundwater pressure-velocity fusion interaction module in a high-precision monitoring system of groundwater velocity and direction based on multi-parameter fusion technology in an embodiment of the present application.

[0023] Figure 4 This is a schematic block diagram of a groundwater spatiotemporal multi-parameter fusion module in a high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology in an embodiment of the present application.

[0024] Figure 5 This is a schematic flowchart of a method for high-precision monitoring of groundwater flow velocity and direction based on multi-parameter fusion technology according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0026] Traditional methods for monitoring groundwater flow often rely on a single indicator, such as groundwater level or flow rate, while ignoring the integrated effects of other related parameters. One disadvantage of this approach is that it does not fully demonstrate the complex nature of the groundwater system, nor does it reveal the interactive effects of different variables. For example, if only the water level is monitored without considering the impact of pressure changes on flow rate or the effect of temperature on rock permeability, critical information may be missed. Moreover, the prediction model constructed based on a single variable often has low prediction accuracy because it fails to cover the interaction between multiple factors, and has weak adaptability when facing the changes in the groundwater system over time, which ultimately weakens the reliability and practical application value of the monitoring work.

[0027] In response to the above technical problems, the technical concept of the present application is to obtain the time sequence of groundwater pressure, the time sequence of groundwater flow velocity and the deep sequence of groundwater temperature values, and use data analysis and processing technology based on deep learning to implicitly associate the groundwater pressure and the groundwater flow velocity in time series, and implicitly associate the groundwater temperature value in space, so as to automatically obtain the short-term prediction value of groundwater flow direction according to the fine-grained matching interaction representation of the time series correlation characteristics of groundwater pressure and the time series correlation characteristics of groundwater flow velocity and the interactive response fusion representation between the spatial implicit correlation characteristics of groundwater temperature values. In this way, by comprehensively analyzing multiple parameters such as groundwater pressure, flow velocity and temperature, the complexity of the groundwater system can be more comprehensively reflected, and the interaction between different parameters can be captured, so as to better understand the dynamic changes of the groundwater system and improve the accuracy of prediction, thereby improving the practicality and intelligence level of the monitoring system.

[0028] Figure 1 This is a schematic block diagram of a high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology in an embodiment of the present application. Figure 2 This is a data flow diagram of a high-precision monitoring system for groundwater flow velocity based on multi-parameter fusion technology in an embodiment of the present application. Figure 1 and Figure 2 As shown, the high-precision monitoring system 100 for groundwater velocity and direction based on multi-parameter fusion technology includes: a groundwater parameter acquisition module 110, which is used to obtain a time sequence of groundwater pressure, a time sequence of groundwater velocity, and a depth sequence of groundwater temperature values; a groundwater parameter implicit association module 120, which is used to extract a groundwater pressure time series implicit association feature vector, a groundwater velocity time series implicit association feature vector, and a groundwater temperature space implicit association feature vector from the time sequence of the groundwater pressure, the time sequence of the groundwater velocity, and the depth sequence of the groundwater temperature values, respectively; a groundwater pressure-velocity fusion module 120, which is used to extract a groundwater pressure time series implicit association feature vector, a groundwater velocity time series implicit association feature vector, and a groundwater temperature space implicit association feature vector from the groundwater pressure time series, the groundwater velocity time series, and the groundwater temperature depth sequence, respectively. The fusion and interaction module 130 is used to perform feature fine-grained optimal fusion interaction on the implicit correlation feature vector of the groundwater pressure time series and the implicit correlation feature vector of the groundwater velocity time series to obtain the groundwater pressure-velocity time series collaborative fusion feature; the groundwater spatiotemporal multi-parameter fusion module 140 is used to perform local and global collaborative fusion interaction on the groundwater pressure-velocity time series collaborative fusion feature and the groundwater temperature space implicit correlation feature vector to obtain the groundwater spatiotemporal multi-parameter fusion representation; the flow direction monitoring result generation module 150 is used to obtain the flow direction monitoring result based on the groundwater spatiotemporal multi-parameter fusion representation.

[0029] In the above-mentioned high-precision monitoring system for groundwater velocity and direction based on multi-parameter fusion technology, the groundwater parameter acquisition module 110 is used to acquire the time sequence of groundwater pressure, the time sequence of groundwater velocity and the depth sequence of groundwater temperature value. In particular, in a specific embodiment of the present application, the time sequence of groundwater pressure is acquired by a pressure sensor, the time sequence of groundwater velocity is acquired by an electromagnetic flowmeter and the depth sequence of groundwater temperature value is acquired by a temperature gradient sensor.

[0030] In the above-mentioned high-precision monitoring system for groundwater velocity and direction based on multi-parameter fusion technology, the groundwater parameter implicit association module 120 is used to extract the groundwater pressure time series implicit association feature vector, the groundwater velocity time series implicit association feature vector and the groundwater temperature space implicit association feature vector from the groundwater pressure time series, the groundwater velocity time series and the groundwater temperature value depth series. It should be understood that considering that the groundwater pressure and the groundwater velocity are constantly changing with the change of time, and there are different degrees of correlation and mutual influence between different time scales, the groundwater temperature value also has different values ​​at different depths, and there are mutual implicit associations between different depth spans. Based on this, in the technical solution of the present application, the groundwater pressure time series implicit association feature vector is extracted from the groundwater pressure time series, the groundwater velocity time series implicit association feature vector is extracted from the groundwater velocity time series, and the groundwater temperature space implicit association feature vector is extracted from the groundwater temperature value depth series. In particular, in a specific embodiment of the present application, the time queue of the groundwater pressure and the time queue of the groundwater flow rate are respectively input into a time series correlation feature extractor based on a bidirectional long short-term memory neural network model to obtain the groundwater pressure time series implicit correlation feature vector and the groundwater flow rate time series implicit correlation feature vector; the depth sequence of the groundwater temperature value is input into a spatial correlation feature extractor based on a one-dimensional convolutional neural network model to obtain the groundwater temperature spatial implicit correlation feature vector.

[0031] In the above-mentioned high-precision monitoring system for groundwater velocity and direction based on multi-parameter fusion technology, the groundwater pressure-velocity fusion interaction module 130 is used to perform feature fine-grained optimal fusion interaction on the implicit correlation feature vector of the groundwater pressure time series and the implicit correlation feature vector of the groundwater velocity time series to obtain the groundwater pressure-velocity time series collaborative fusion feature. It should be understood that considering the local temporal correlation relationship between the implicit correlation feature of the groundwater pressure time series and the implicit correlation feature of the groundwater velocity time series, when the groundwater pressure in a certain area increases at a certain time, it may cause the flow velocity near the area to accelerate, because the groundwater flows from the high-pressure area to the low-pressure area to form a flow. Therefore, in order to analyze and understand the mutual correlation and interaction between the two in a more fine-grained manner, so that some features with very weak correlation or imperceptible can be more prominent, in the technical solution of the present application, the implicit correlation feature vector of the groundwater pressure time series and the implicit correlation feature vector of the groundwater flow velocity time series are subjected to fine-grained optimal fusion interaction to obtain the groundwater pressure-flow velocity time series collaborative alignment fusion feature vector as the groundwater pressure-flow velocity time series collaborative fusion feature.

[0032] Optionally, in one embodiment of the present application, Figure 3 This is a schematic block diagram of a groundwater pressure-velocity fusion interaction module in a groundwater velocity and direction high-precision monitoring system based on multi-parameter fusion technology in an embodiment of the present application. Figure 3 As shown, the groundwater pressure-flow velocity fusion interaction module 130 includes: a groundwater pressure-flow velocity time series fine-grained feature decomposition unit 131, which is used to perform fine-grained feature decomposition on the groundwater pressure time series implicit correlation feature vector and the groundwater flow velocity time series implicit correlation feature vector based on a predetermined scale to obtain a set of groundwater pressure time series implicit correlation sub-component feature vectors and a set of groundwater flow velocity time series implicit correlation sub-component feature vectors; a groundwater pressure-flow velocity query matching fusion unit 132, which is used to perform query matching alignment fusion on the set of groundwater pressure time series implicit correlation sub-component feature vectors and the set of groundwater flow velocity time series implicit correlation sub-component feature vectors to obtain a groundwater pressure-flow velocity time series collaborative alignment fusion feature vector as the groundwater pressure-flow velocity time series collaborative fusion feature.

[0033] Optionally, in one embodiment of the present application, the groundwater pressure and flow velocity query matching fusion unit includes: a groundwater pressure-flow velocity time series measurement subunit, which is used to calculate the groundwater pressure-flow velocity time series hyperbolic space distance measurement factor between any two groundwater pressure time series implicitly associated subcomponent feature vectors and the groundwater flow velocity time series implicitly associated subcomponent feature vectors in the set of groundwater pressure time series implicitly associated subcomponent feature vectors and the set of groundwater flow velocity time series implicitly associated subcomponent feature vectors; and an optimal characteristic subcomponent pairing screening subunit, which is used to perform optimal characteristic subcomponent pairing screening on the set of groundwater pressure time series implicitly associated subcomponent feature vectors and the set of groundwater flow velocity time series implicitly associated subcomponent feature vectors based on the groundwater pressure-flow velocity time series hyperbolic space distance measurement factor to obtain the optimal characteristic subcomponent pairing. A set of optimal {groundwater pressure time series implicitly associated sub-component feature vectors, groundwater flow velocity time series implicitly associated sub-component feature vectors} pairs; a groundwater pressure-flow velocity multi-dimensional interaction sub-unit, used to input each optimal {groundwater pressure time series implicitly associated sub-component feature vectors, groundwater flow velocity time series implicitly associated sub-component feature vectors} pair in the set of optimal {groundwater pressure time series implicitly associated sub-component feature vectors} pairs into a feature multi-dimensional interaction module to obtain a set of groundwater pressure-flow velocity optimal sub-component feature pairing interactive fusion feature vectors; a feature vector cascade sub-unit, used to cascade the set of groundwater pressure-flow velocity optimal sub-component feature pairing interactive fusion feature vectors to obtain the groundwater pressure-flow velocity time series collaborative alignment fusion feature vector.

[0034] Optionally, in one embodiment of the present application, the groundwater pressure-flow velocity time series measurement subunit is used to: extract a predetermined groundwater pressure time series implicit correlation subcomponent feature vector from the set of groundwater pressure time series implicit correlation subcomponent feature vectors; calculate the hyperbolic space distance measurement factor between the predetermined groundwater pressure time series implicit correlation subcomponent feature vector and each groundwater flow velocity time series implicit correlation subcomponent feature vector in the set of groundwater flow velocity time series implicit correlation subcomponent feature vectors to obtain a set of groundwater pressure-flow velocity time series hyperbolic space distance measurement factors of the predetermined groundwater pressure time series implicit correlation subcomponent feature vector.

[0035] Optionally, in one embodiment of the present application, the hyperbolic space distance measurement factor between each groundwater velocity time series implicit correlation subcomponent feature vector in the set of the predetermined groundwater pressure time series implicit correlation subcomponent feature vector and the groundwater velocity time series implicit correlation subcomponent feature vector is calculated to obtain a set of groundwater pressure-velocity time series hyperbolic space distance measurement factors of the predetermined groundwater pressure time series implicit correlation subcomponent feature vector, including: calculating the position difference between the predetermined groundwater pressure time series implicit correlation subcomponent feature vector and the groundwater velocity time series implicit correlation subcomponent feature vector to obtain a groundwater pressure-velocity time series differential vector, and calculating the square of the norm of the groundwater pressure-velocity time series differential vector to obtain A groundwater pressure-velocity time series difference norm value; respectively calculating the difference between a constant one and the square of a norm of a predetermined groundwater pressure time series implicitly associated subcomponent eigenvector and a groundwater velocity time series implicitly associated subcomponent eigenvector to obtain a predetermined groundwater pressure time series difference value and a groundwater velocity time series difference value; dividing the groundwater pressure-velocity time series difference norm value by the product value between the predetermined groundwater pressure time series difference value and the groundwater velocity time series difference value, multiplying the obtained value by a constant two and then adding the value to a constant one to obtain a groundwater pressure-velocity time series distance measurement value; inputting the groundwater pressure-velocity time series distance measurement value into an arccosh function to obtain a groundwater pressure-velocity time series hyperbolic space distance measurement factor.

[0036] Optionally, in one embodiment of the present application, the groundwater pressure-flow velocity multidimensional interaction subunit is used to: respectively calculate the position difference, position dot product and position addition between the optimal {groundwater pressure time series implicit correlation subcomponent feature vector, groundwater flow velocity time series implicit correlation subcomponent feature vector} pairings to obtain a groundwater pressure-flow velocity time series difference vector, a groundwater pressure-flow velocity time series dot product vector and a groundwater pressure-flow velocity time series addition vector; and add the groundwater pressure-flow velocity time series difference vector, the groundwater pressure-flow velocity time series dot product vector and the groundwater pressure-flow velocity time series addition vector. The groundwater pressure-velocity time series sum vector is multiplied by the first weighted parameter, the second weighted parameter and the third weighted parameter respectively by position to obtain the groundwater pressure-velocity time series difference weighted vector, the groundwater pressure-velocity time series point product weighted vector and the groundwater pressure-velocity time series sum weighted vector; the groundwater pressure-velocity time series difference weighted vector, the groundwater pressure-velocity time series point product weighted vector and the groundwater pressure-velocity time series sum weighted vector are calculated by position point to obtain the groundwater pressure-velocity optimal sub-component feature pairing interactive fusion feature vector.

[0037] Optionally, in one embodiment of the present application, based on the groundwater pressure-flow velocity time series hyperbolic space distance measurement factor, the set of groundwater pressure time series implicitly associated sub-component feature vectors and the set of groundwater flow velocity time series implicitly associated sub-component feature vectors are screened for the best feature sub-component pairing to obtain the set of optimal {groundwater pressure time series implicitly associated sub-component feature vectors, groundwater flow velocity time series implicitly associated sub-component feature vectors} pairings, including: taking the groundwater flow velocity time series implicitly associated sub-component feature vector and the predetermined groundwater pressure time series implicitly associated sub-component feature vector corresponding to the smallest groundwater pressure-flow velocity time series hyperbolic space distance measurement factor in the set of groundwater pressure-flow velocity time series hyperbolic space distance measurement factors of the predetermined groundwater pressure time series implicitly associated sub-component feature vectors as the optimal {groundwater pressure time series implicitly associated sub-component feature vector, groundwater flow velocity time series implicitly associated sub-component feature vector} pairing.

[0038] In summary, in the embodiment of the present application, the implicit correlation feature vector of the groundwater pressure time series and the implicit correlation feature vector of the groundwater velocity time series are subjected to the optimal fusion interaction of fine-grained features, which can be expressed by the formula:

[0039] Decouple(v1)={v 11 ,v 12 ,...,v 1i ,...,v 1n}

[0040] Decouple(v2) = {v 21 ,v22 ,...,v 2j ,...,v 2n}

[0041]

[0042]

[0043] f1=[v p1 ;v p2 ;...;v pn ]

[0044] Among them, v1 is the implicit correlation eigenvector of the groundwater pressure time series, v 11 ,v 12 ,...,v 1i ,...,v 1n are the first, second, ..., i-th, ..., n-th groundwater pressure time series implicit correlation sub-component feature vectors in the set of the groundwater pressure time series implicit correlation sub-component feature vectors, v2 is the groundwater velocity time series implicit correlation feature vector, v 21 ,v 22 ,...,v 2j ,...,v 2n are the first, second, ..., jth, ..., nth groundwater velocity time series implicitly associated sub-component feature vectors in the set of the groundwater velocity time series implicitly associated sub-component feature vectors, Decouple(·) is a fine-grained feature decomposition operation, d P (v 1i ,v 2j ) is used to calculate v 1i and v 2j The distance measurement factor between the groundwater pressure and velocity time series in hyperbolic space, arg min j (·) is the j value corresponding to the minimum value returned, k is the minimum approximate matching value, ‖·‖ 2 To calculate the square of the one-norm of a vector, arccos(·) is the inverse hyperbolic cosine function, ⊙ and are position difference, position point multiplication and position addition respectively, δ, ε and γ are the corresponding weighting parameters, v 2k is the kth groundwater velocity time series implicitly associated subcomponent feature vector in the set of groundwater velocity time series implicitly associated subcomponent feature vectors, v pi is the set of the groundwater pressure-flow velocity optimal sub-component feature pairing interactive fusion feature vectors, the ith groundwater pressure-flow velocity optimal sub-component feature pairing interactive fusion feature vector, v p1 、v p2 and vpn They are respectively the first, second and nth groundwater pressure-velocity optimal sub-component feature pairing interactive fusion feature vectors in the set of groundwater pressure-velocity optimal sub-component feature pairing interactive fusion feature vectors, [·; ·; ...; ·] is cascade processing, and f1 is the groundwater pressure-velocity time series collaborative alignment fusion feature vector.

[0045] Specifically, firstly, the implicit correlation feature vector of the groundwater pressure time series and the implicit correlation feature vector of the groundwater velocity time series are fine-grained feature decomposition based on a predetermined scale to decompose them into a set of implicit correlation sub-component feature vectors of the groundwater pressure time series and a set of implicit correlation sub-component feature vectors of the groundwater velocity time series containing multiple local time series, which provides a basis for the subsequent fine-grained optimal interactive matching. Then, the hyperbolic space distance measurement factor between each implicit correlation sub-component feature of the groundwater pressure time series and the implicit correlation sub-component feature of the groundwater velocity time series in the two sub-component sets is calculated to obtain a sequence of hyperbolic space distance measurement factors of the groundwater pressure-velocity time series. That is, by calculating the hyperbolic space distance measurement factor, the similarity between the groundwater pressure sub-component feature vector and the groundwater velocity sub-component feature vector can be quantified. The smaller the distance, the higher the similarity between the two, which is the basis for the subsequent fine-grained optimal interactive matching. In particular, due to the complexity and hierarchy of the groundwater system, the hyperbolic space distance metric factor utilizes the characteristics of hyperbolic geometry. This geometric structure is more suitable for processing data with hierarchical structure or long-tail distribution. Compared with the traditional Euclidean distance, the distance metric factor of the hyperbolic space can better capture the nonlinear relationship and long-distance dependence between data. Afterwards, the optimal {groundwater pressure time series implicit correlation subcomponent feature vector, groundwater flow velocity time series implicit correlation subcomponent feature vector} pairing set is screened based on the metric factor to ensure that the feature vectors of groundwater pressure and flow velocity are correctly aligned in time to avoid errors caused by time misalignment, and to ensure that each pairing relationship can reflect the similarity between the two sets of feature vectors to the greatest extent, so as to maximize the retention of the most critical data in the subsequent information integration stage. Furthermore, in order to further capture and refine the complex interactive relationship between the best pairs, these best pairs are input into the feature multi-dimensional interaction module to obtain the set of groundwater pressure-flow velocity optimal subcomponent feature pairing interactive fusion feature vectors. Finally, the set of fused feature vectors is cascaded to obtain the final groundwater pressure-velocity time series co-aligned fused feature vector.

[0046] In the above-mentioned high-precision monitoring system for groundwater velocity and direction based on multi-parameter fusion technology, the groundwater spatiotemporal multi-parameter fusion module 140 is used to perform local and global collaborative fusion interaction on the groundwater pressure-velocity time series collaborative fusion characteristics and the groundwater temperature space implicit correlation feature vector to obtain a groundwater spatiotemporal multi-parameter fusion representation. It should be understood that in order to fuse the groundwater pressure-velocity time series collaborative alignment fusion feature vector and the groundwater temperature space implicit correlation feature vector to generate a comprehensive groundwater global representation, thereby more accurately predicting the groundwater flow direction and other related parameters, in the technical solution of the present application, the groundwater pressure-velocity time series collaborative fusion characteristics and the groundwater temperature space implicit correlation feature vector are locally and globally collaboratively fused to obtain a groundwater spatiotemporal multi-parameter fusion representation vector.

[0047] Optionally, in one embodiment of the present application, Figure 4 Schematic block diagram of a groundwater spatiotemporal multi-parameter fusion module in a groundwater velocity and direction high-precision monitoring system based on multi-parameter fusion technology in an embodiment of the present application. Figure 4 As shown, the groundwater spatiotemporal multi-parameter fusion module 140 includes: a groundwater temperature decoupling unit 141, which is used to perform feature decoupling on the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of groundwater temperature local spatial implicit correlation features; a groundwater multi-parameter interactive fusion unit 142, which is used to perform feature interactive fusion on the groundwater pressure-velocity time series collaborative fusion feature and the sequence of groundwater temperature local spatial implicit correlation features to obtain the groundwater spatiotemporal multi-parameter fusion representation.

[0048] Optionally, in one embodiment of the present application, the groundwater temperature decoupling unit is used to: perform feature decoupling on the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of groundwater temperature local spatial implicit correlation feature vectors as a sequence of groundwater temperature local spatial implicit correlation features.

[0049] Optionally, in one embodiment of the present application, the groundwater multi-parameter interactive fusion unit includes: a nonlinear transformation processing subunit, which is used to perform a nonlinear transformation on the groundwater pressure-velocity time series collaborative alignment fusion feature vector to obtain the groundwater pressure-velocity time series collaborative alignment fusion feature vector after the nonlinear transformation; an association interaction processing subunit, which is used to calculate the association interaction vector between the groundwater pressure-velocity time series collaborative alignment fusion feature vector after the nonlinear transformation and each groundwater temperature local space implicit association feature vector in the sequence of groundwater temperature local space implicit association feature vectors to obtain a sequence of groundwater spatiotemporal multi-parameter association interaction vectors; a groundwater spatiotemporal multi-parameter feature interaction energy factor calculation subunit, which is used to calculate the The characteristic interaction energy factors of each groundwater spatiotemporal multi-parameter associated interaction vector in the sequence of groundwater spatiotemporal multi-parameter associated interaction vectors are used to obtain a sequence of groundwater spatiotemporal multi-parameter characteristic interaction energy factors; a normalization processing subunit is used to input the sequence of groundwater spatiotemporal multi-parameter characteristic interaction energy factors into a Sigmoid function to obtain a sequence of normalized groundwater spatiotemporal multi-parameter characteristic interaction energy factors; a groundwater spatiotemporal multi-parameter fusion processing subunit is used to use the sequence of normalized groundwater spatiotemporal multi-parameter characteristic interaction energy factors as a sequence of weights, calculate the position-weighted sum of the sequence of groundwater spatiotemporal multi-parameter associated interaction vectors to obtain a groundwater spatiotemporal multi-parameter fusion representation vector as the groundwater spatiotemporal multi-parameter fusion representation.

[0050] Optionally, in one embodiment of the present application, the correlation interaction vectors between the groundwater pressure-flow velocity time series coordinated alignment fusion feature vector after the nonlinear transformation and each groundwater temperature local space implicit correlation feature vector in the sequence of the groundwater temperature local space implicit correlation feature vector are calculated to obtain a sequence of groundwater spatiotemporal multi-parameter correlation interaction vectors, including: respectively calculating the positional subtraction, positional dot multiplication and positional multiplication between the groundwater pressure-flow velocity time series coordinated alignment fusion feature vector after the nonlinear transformation and the groundwater temperature local space implicit correlation feature vector. The groundwater spatiotemporal multi-parameters are added in positions to obtain a groundwater spatiotemporal multi-parameter difference vector, a groundwater spatiotemporal multi-parameter dot product vector and a groundwater spatiotemporal multi-parameter sum vector; the groundwater spatiotemporal multi-parameter difference vector, the groundwater spatiotemporal multi-parameter dot product vector and the groundwater spatiotemporal multi-parameter sum vector are cascaded to obtain a groundwater spatiotemporal multi-parameter cascade vector; after performing one-dimensional convolution coding on the groundwater spatiotemporal multi-parameter cascade vector, the obtained groundwater spatiotemporal multi-parameter convolution coding vector is subjected to maximum pooling processing based on a local window to obtain the groundwater spatiotemporal multi-parameter association interaction vector.

[0051] Optionally, in one embodiment of the present application, the groundwater spatiotemporal multi-parameter characteristic interaction energy factor calculation subunit is used to: respectively calculate the mean and variance of the groundwater spatiotemporal multi-parameter correlation interaction vector to obtain the groundwater spatiotemporal multi-parameter correlation interaction mean and the groundwater spatiotemporal multi-parameter correlation interaction variance; calculate the positional difference between the groundwater spatiotemporal multi-parameter correlation interaction vector and the groundwater spatiotemporal multi-parameter correlation interaction mean to obtain the groundwater spatiotemporal multi-parameter correlation interaction difference vector; calculate the fourth power of each eigenvalue in the groundwater spatiotemporal multi-parameter correlation interaction difference vector to obtain the groundwater spatiotemporal multi-parameter correlation interaction modulation difference vector; calculate the expected value of the groundwater spatiotemporal multi-parameter correlation interaction modulation difference vector to obtain the groundwater spatiotemporal multi-parameter correlation interaction expected value; compare the groundwater spatiotemporal multi-parameter correlation interaction expected value with the square of the groundwater spatiotemporal multi-parameter correlation interaction variance. Divide to obtain the groundwater spatiotemporal multi-parameter association interaction kurtosis value; calculate the sum of the groundwater spatiotemporal multi-parameter association interaction variance and the hyperparameter to obtain the groundwater spatiotemporal multi-parameter association interaction first energy factor; calculate the square of the difference between the groundwater spatiotemporal multi-parameter association interaction kurtosis value and the groundwater spatiotemporal multi-parameter association interaction mean to obtain the groundwater spatiotemporal multi-parameter association interaction difference value; add the value obtained by multiplying the groundwater spatiotemporal multi-parameter association interaction variance by constant two and the value obtained by multiplying the hyperparameter by constant two to the groundwater spatiotemporal multi-parameter association interaction difference value to obtain the groundwater spatiotemporal multi-parameter association interaction second energy factor; divide the groundwater spatiotemporal multi-parameter association interaction first energy factor by the groundwater spatiotemporal multi-parameter association interaction second energy factor to obtain the groundwater spatiotemporal multi-parameter characteristic interaction energy factor corresponding to the groundwater spatiotemporal multi-parameter association interaction vector.

[0052] In summary, in the embodiment of the present application, the groundwater pressure-velocity time series coordinated alignment fusion feature vector and the groundwater temperature spatial implicit correlation feature vector are locally and globally coordinated and fused, which can be expressed by the formula:

[0053] Decouple(f2)={f 21 ,f 22 ,...,f 2i ,...,f 2n}

[0054] f 1t =Wf1+b

[0055]

[0056] E norm,12i =sigmoid(E 12i )

[0057] v m =∑i f 12i ⊙E norm,12i

[0058] Where f2 is the implicit correlation eigenvector of the groundwater temperature space, Decouple(·) is the characteristic decoupling operation, and f 21 ,f 22 ,...,f 2i ,...,f 2n are the 1st, 2nd, ..., ith, ..., nth groundwater temperature local space implicit correlation feature vectors in the sequence of the groundwater temperature local space implicit correlation feature vectors, n is the number of feature vectors in the sequence of the groundwater temperature local space implicit correlation feature vectors, f1 is the groundwater pressure-velocity time series collaborative alignment fusion feature vector, W and b are the transformation matrix and the bias vector respectively, f 1t The groundwater pressure-velocity time series collaborative alignment fusion feature vector after the nonlinear transformation is: ⊙ and are position-wise subtraction, position-wise multiplication, and position-wise addition, respectively. [·,·,·] is a cascade process. conv1D(·) is a one-dimensional convolutional encoding operation. MaxPool(·) is a maximum pooling operation. 12i is the groundwater spatiotemporal multi-parameter correlation interaction vector between the local spatial implicit correlation feature vector of groundwater temperature and the groundwater pressure-velocity time series coordinated alignment fusion feature vector after the nonlinear transformation, f 12i (t) is f 12i The eigenvalues ​​at each position in 12i Yes 12i The mean of , E(·) is the expected value, σ 12i 2 Yes 12i The variance of k 12i f 12i The kurtosis value, λ is a hyperparameter, E 12i Yes 12i The corresponding groundwater spatiotemporal multi-parameter characteristic interaction energy factor, sigmoid(·) is the sigmoid function, E norm,12i It is M 12i The corresponding normalized groundwater spatiotemporal multi-parameter characteristic interaction energy factor, ⊙ is the point multiplication by location, v m is the groundwater spatiotemporal multi-parameter fusion representation vector.

[0059] In detail, first, the sequence of local spatial implicit correlation feature vectors of groundwater temperature is obtained by feature decoupling the implicit correlation feature vector of groundwater temperature to reveal the features of different spatial dimensions within the feature vector. Then, the groundwater pressure-velocity time series coordinated alignment fusion feature vector is nonlinearly changed to explain the more complex relationship between groundwater pressure and flow, and the groundwater pressure-velocity time series coordinated alignment fusion feature vector after nonlinear transformation is obtained. Then, in order to analyze and understand the local correlation interaction relationship between various groundwater parameters in a more fine-grained manner, the nonlinearly transformed features are associated and interacted with each local spatial implicit correlation feature of groundwater temperature to obtain a sequence of groundwater spatiotemporal multi-parameter correlation interaction vectors. Subsequently, the feature interaction energy factor is extracted from each correlation interaction vector to obtain a sequence of groundwater spatiotemporal multi-parameter feature interaction energy factors. That is, the feature interaction energy factor reveals the closeness and significance of the interaction between feature vectors, and extracting this factor plays an important role in discovering and enhancing those feature relationships that are critical to the model prediction performance. Furthermore, in order to convert these factors into weights, each energy factor is input into the Sigmoid function for normalization so that its value range falls within the interval (0,1), so that the model can more reasonably allocate the attention to the interaction of different features, and obtain the sequence of normalized groundwater spatiotemporal multi-parameter feature interaction energy factors. Finally, based on the sequence of normalized factors, the sequence of groundwater spatiotemporal multi-parameter correlation interaction vectors is weighted fused to integrate the interaction information of local features, and a comprehensive and integrated groundwater spatiotemporal multi-parameter fusion representation vector is obtained.

[0060] In the above-mentioned high-precision monitoring system for groundwater velocity and direction based on multi-parameter fusion technology, the flow direction monitoring result generation module 150 is used to obtain the flow direction monitoring result based on the groundwater spatiotemporal multi-parameter fusion representation. Optionally, in one embodiment of the present application, the flow direction monitoring result generation module is used to: input the groundwater spatiotemporal multi-parameter fusion representation vector into the groundwater flow direction monitoring module based on the decoder to obtain the flow direction monitoring result, and the flow direction monitoring result is the short-term prediction decoding value of the groundwater flow direction. That is, the groundwater spatiotemporal multi-parameter fusion representation obtained by local-global fusion of the groundwater pressure-velocity time series collaborative alignment fusion feature vector and the groundwater temperature space implicit correlation feature vector is decoded to automatically obtain the short-term prediction value of the groundwater flow direction. In this way, by comprehensively analyzing the groundwater pressure, velocity and temperature, the complexity of the groundwater system can be more comprehensively reflected, and the interaction between different parameters can be captured, so as to better understand the dynamic changes of the groundwater system and improve the accuracy of the prediction, thereby improving the practicality and intelligence level of the monitoring system.

[0061] In a preferred example, the groundwater spatiotemporal multi-parameter fusion representation vector is passed through a decoder-based groundwater flow direction monitoring module to obtain a flow direction monitoring result, including:

[0062] Calculate the sum of the absolute values ​​of each eigenvalue of the groundwater spatiotemporal multi-parameter fusion representation vector to obtain a first groundwater spatiotemporal multi-parameter fusion representation and modulation value, and calculate the square root of the sum of the squares of each eigenvalue of the groundwater spatiotemporal multi-parameter fusion representation vector to obtain a second groundwater spatiotemporal multi-parameter fusion representation and modulation value;

[0063] After point-wise subtraction of the groundwater spatiotemporal multi-parameter fusion representation vector and the second groundwater spatiotemporal multi-parameter fusion representation and modulation value, point-wise multiplication is performed with the number of eigenvalues ​​of the groundwater spatiotemporal multi-parameter fusion representation vector and the reciprocal of the first groundwater spatiotemporal multi-parameter fusion representation and modulation value, and the reciprocal of each eigenvalue is taken to obtain the first groundwater spatiotemporal multi-parameter fusion representation phase conversion vector;

[0064] After point-wise subtraction of the groundwater spatiotemporal multi-parameter fusion representation vector and the first groundwater spatiotemporal multi-parameter fusion representation and modulation value, point-wise multiplication is performed with the square root of the number of eigenvalues ​​of the groundwater spatiotemporal multi-parameter fusion representation vector and the reciprocal of the second groundwater spatiotemporal multi-parameter fusion representation and modulation value, and the reciprocal of each eigenvalue is taken to obtain the second groundwater spatiotemporal multi-parameter fusion representation phase conversion vector; and

[0065] Subtract the dot product vector of the second groundwater spatiotemporal multi-parameter fusion representation phase conversion vector and the weighted hyperparameter from the first groundwater spatiotemporal multi-parameter fusion representation phase conversion vector to obtain an optimized groundwater spatiotemporal multi-parameter fusion representation vector; and

[0066] The optimized groundwater spatiotemporal multi-parameter fusion representation vector is passed through a decoder-based groundwater flow direction monitoring module to obtain a flow direction monitoring result.

[0067] Here, the groundwater spatiotemporal multi-parameter fusion representation vector, denoted as V, is optimally represented as:

[0068]

[0069] v i ∈V∈R n

[0070] Where V is the groundwater spatiotemporal multi-parameter fusion representation vector, R represents a real number set, and v irepresents the eigenvalue of the i-th position of the groundwater spatiotemporal multi-parameter fusion representation vector, n represents the number of eigenvalues ​​of the groundwater spatiotemporal multi-parameter fusion representation vector, α represents the first groundwater spatiotemporal multi-parameter fusion representation and modulation value, β represents the second groundwater spatiotemporal multi-parameter fusion representation and modulation value, ⊙ represents point multiplication by position, θ represents subtraction by position, [·] ⊙-1 is the inverse of each eigenvalue in the vector, V1 represents the first groundwater spatiotemporal multi-parameter fusion representation phase transition vector, V2 represents the second groundwater spatiotemporal multi-parameter fusion representation phase transition vector, ω represents the weighted hyperparameter, and V′ represents the optimized groundwater spatiotemporal multi-parameter fusion representation vector.

[0071] Here, in the technical solution of the present application, the groundwater pressure-velocity time series collaborative alignment fusion feature vector and the groundwater temperature space implicit correlation feature vector respectively represent the groundwater pressure-velocity time series collaborative fusion feature and the groundwater temperature space implicit correlation feature, so that when they are subjected to local and global collaborative fusion interaction, the time series correlation feature distribution pattern based on the source time series distribution will also have local-global time series interactive response differences. Therefore, it is expected to improve the detailed interactive response semantic aggregation expression effect of the groundwater spatiotemporal multi-parameter fusion representation vector based on the association pattern representation difference.

[0072] Therefore, in the preferred example, the difference of the characteristic value of the groundwater spatiotemporal multi-parameter fusion representation vector relative to the vector feature set of the groundwater spatiotemporal multi-parameter fusion representation vector as a whole and the difference of the modulation representation is used as the semantic change intensity information, and the phase-like conversion corresponding to the position-based intensity modulation is performed through different and modulation representation forms, so that the aggregation enhancement of the semantic change phase perception can improve the axial aggregation receptive field along the feature aggregation direction by performing a spatial translation operation based on alternating stacking under the vector set scale balance of the groundwater spatiotemporal multi-parameter fusion representation vector, thereby improving the perception effect of the aggregation semantics of the groundwater spatiotemporal multi-parameter fusion representation vector for the semantic change of details, so as to improve the expression effect of the groundwater spatiotemporal multi-parameter fusion representation vector, and improve the accuracy of the flow direction monitoring result obtained by the groundwater flow direction monitoring module based on the decoder. In this way, by comprehensively analyzing the groundwater pressure, flow velocity and temperature, the complexity of the groundwater system can be more comprehensively reflected, and the interaction between different parameters can be captured, so as to better understand the dynamic changes of the groundwater system and improve the accuracy of the prediction, thereby improving the practicality and intelligence level of the monitoring system.

[0073] In summary, according to the embodiment of the present application, a high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology is explained, which obtains a through-hole filling state image collected by an industrial camera, obtains the time queue of groundwater pressure, the time queue of groundwater flow velocity and the depth sequence of groundwater temperature value, and adopts data analysis and processing technology based on deep learning to implicitly associate the groundwater pressure and the groundwater flow velocity in time series, and implicitly associate the groundwater temperature value in space, so as to automatically obtain the short-term prediction value of groundwater flow direction according to the fine-grained matching interaction representation of the time series correlation characteristics of groundwater pressure and the time series correlation characteristics of groundwater flow velocity and the interactive response fusion representation between the spatial implicit correlation characteristics of groundwater temperature value. In this way, by comprehensively analyzing multiple parameters such as groundwater pressure, flow velocity and temperature, the complexity of the groundwater system can be more comprehensively reflected, and the interaction between different parameters can be captured, so as to better understand the dynamic changes of the groundwater system and improve the accuracy of prediction, thereby improving the practicality and intelligence level of the monitoring system.

[0074] Figure 5 Schematic flow chart of a method for high-precision monitoring of groundwater flow velocity and direction based on multi-parameter fusion technology according to an embodiment of the present application. Figure 5 As shown, a high-precision monitoring method for groundwater velocity and direction based on multi-parameter fusion technology includes: S1, obtaining a time queue of groundwater pressure, a time queue of groundwater velocity and a depth sequence of groundwater temperature values; S2, extracting a groundwater pressure time series implicit correlation feature vector, a groundwater velocity time series implicit correlation feature vector and a groundwater temperature space implicit correlation feature vector from the time queue of groundwater pressure, the time queue of groundwater velocity and the depth sequence of groundwater temperature values ​​respectively; S3, performing fine-grained optimal fusion interaction on the groundwater pressure time series implicit correlation feature vector and the groundwater velocity time series implicit correlation feature vector to obtain a groundwater pressure-velocity time series collaborative fusion feature; S4, performing local and global collaborative fusion interaction on the groundwater pressure-velocity time series collaborative fusion feature and the groundwater temperature space implicit correlation feature vector to obtain a groundwater spatiotemporal multi-parameter fusion representation; S5, obtaining a flow direction monitoring result based on the groundwater spatiotemporal multi-parameter fusion representation.

[0075] Optionally, in one embodiment of the present application, the groundwater pressure-velocity time series collaborative fusion characteristics and the groundwater temperature spatial implicit correlation feature vector are subjected to local and global collaborative fusion interaction, including: feature decoupling of the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of local spatial implicit correlation characteristics of groundwater temperature; feature interactive fusion of the groundwater pressure-velocity time series collaborative fusion characteristics and the sequence of local spatial implicit correlation characteristics of groundwater temperature to obtain the groundwater spatiotemporal multi-parameter fusion representation.

[0076] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0077] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the module division is only a logical function division, and there may be other division methods in actual implementation. 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 units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0078] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0079] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0080] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units stated in the device claim can also be implemented by one unit through software or hardware.

[0081] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology, characterized in that: include: A groundwater parameter acquisition module is used to obtain a time sequence of groundwater pressure, a time sequence of groundwater velocity, and a depth sequence of groundwater temperature values; A groundwater parameter implicit association module is used to extract a groundwater pressure time series implicit association feature vector, a groundwater flow velocity time series implicit association feature vector and a groundwater temperature space implicit association feature vector from the groundwater pressure time series, the groundwater flow velocity time series and the groundwater temperature value depth sequence respectively; A groundwater pressure-velocity fusion interaction module is used to perform fine-grained optimal fusion interaction on the groundwater pressure time series implicit correlation feature vector and the groundwater velocity time series implicit correlation feature vector to obtain a groundwater pressure-velocity time series collaborative fusion feature; A groundwater spatiotemporal multi-parameter fusion module, used for performing local and global collaborative fusion interaction on the groundwater pressure-velocity time series collaborative fusion characteristics and the groundwater temperature spatial implicit correlation feature vector to obtain a groundwater spatiotemporal multi-parameter fusion representation, wherein the groundwater spatiotemporal multi-parameter fusion module includes: a groundwater temperature decoupling unit, used for performing feature decoupling on the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of groundwater temperature local spatial implicit correlation characteristics; a groundwater multi-parameter interactive fusion unit, used for performing feature interactive fusion on the groundwater pressure-velocity time series collaborative fusion characteristics and the sequence of groundwater temperature local spatial implicit correlation characteristics to obtain the groundwater spatiotemporal multi-parameter fusion representation; The flow direction monitoring result generating module is used to obtain the flow direction monitoring result based on the groundwater spatiotemporal multi-parameter fusion representation.

2. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 1 is characterized in that: The groundwater pressure-velocity fusion interaction module includes: A groundwater pressure and velocity time series fine-grained feature decomposition unit, used for performing fine-grained feature decomposition on the groundwater pressure time series implicit correlation feature vector and the groundwater velocity time series implicit correlation feature vector based on a predetermined scale to obtain a set of groundwater pressure time series implicit correlation sub-component feature vectors and a set of groundwater velocity time series implicit correlation sub-component feature vectors; The groundwater pressure and velocity query matching and fusion unit is used to query, match, align and fuse the set of implicitly correlated sub-component feature vectors of the groundwater pressure time series and the set of implicitly correlated sub-component feature vectors of the groundwater velocity time series to obtain a groundwater pressure-velocity time series collaborative alignment fusion feature vector as the groundwater pressure-velocity time series collaborative fusion feature.

3. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 2 is characterized in that: The groundwater pressure and velocity query matching and fusion unit comprises: A groundwater pressure-velocity time series measurement subunit, used to calculate a groundwater pressure-velocity time series hyperbolic space distance measurement factor between any two groundwater pressure time series implicit correlation subcomponent feature vectors and groundwater velocity time series implicit correlation subcomponent feature vectors in the set of groundwater pressure time series implicit correlation subcomponent feature vectors and the set of groundwater velocity time series implicit correlation subcomponent feature vectors; The optimal characteristic sub-component pairing screening sub-unit is used to perform optimal characteristic sub-component pairing screening on the set of the groundwater pressure time series implicit correlation sub-component feature vectors and the set of the groundwater flow velocity time series implicit correlation sub-component feature vectors based on the groundwater pressure-flow velocity time series hyperbolic space distance measurement factor to obtain the optimal {groundwater pressure time series implicit correlation sub-component feature vector, groundwater flow velocity time series implicit correlation sub-component feature vector} pairing set; A groundwater pressure-flow velocity multi-dimensional interaction sub-unit, used for inputting each optimal {groundwater pressure time series implicit correlation sub-component feature vector, groundwater flow velocity time series implicit correlation sub-component feature vector} pair in the set of the optimal {groundwater pressure time series implicit correlation sub-component feature vector, groundwater flow velocity time series implicit correlation sub-component feature vector} pair into a feature multi-dimensional interaction module to obtain a set of groundwater pressure-flow velocity optimal sub-component feature pairing interaction fusion feature vectors; The feature vector cascade subunit is used to cascade the set of the groundwater pressure-flow velocity optimal sub-component feature pairing interactive fusion feature vectors to obtain the groundwater pressure-flow velocity time series collaborative alignment fusion feature vector.

4. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 3 is characterized in that: The groundwater pressure-flow velocity time series measurement subunit is used for: Extracting a predetermined groundwater pressure time series implicit correlation subcomponent feature vector from the set of groundwater pressure time series implicit correlation subcomponent feature vectors; The hyperbolic space distance measurement factor between the predetermined groundwater pressure time series implicit correlation sub-component feature vector and each groundwater velocity time series implicit correlation sub-component feature vector in the set of the groundwater velocity time series implicit correlation sub-component feature vector is calculated to obtain a set of groundwater pressure-velocity time series hyperbolic space distance measurement factors of the predetermined groundwater pressure time series implicit correlation sub-component feature vector.

5. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 4 is characterized in that: The groundwater pressure-velocity multi-dimensional interaction subunit is used to: Respectively calculating the position difference, position dot product and position addition between the optimal {groundwater pressure time series implicit correlation subcomponent eigenvector, groundwater velocity time series implicit correlation subcomponent eigenvector} pairs to obtain a groundwater pressure-velocity time series difference vector, a groundwater pressure-velocity time series dot product vector and a groundwater pressure-velocity time series sum vector; The groundwater pressure-velocity time series difference vector, the groundwater pressure-velocity time series point product vector and the groundwater pressure-velocity time series sum vector are respectively multiplied by position with the first weighting parameter, the second weighting parameter and the third weighting parameter to obtain a groundwater pressure-velocity time series difference weighted vector, a groundwater pressure-velocity time series point product weighted vector and a groundwater pressure-velocity time series sum weighted vector; The groundwater pressure-velocity optimal sub-component feature pairing interactive fusion feature vector of the groundwater pressure-velocity time series difference weighted vector, the groundwater pressure-velocity time series point multiplication weighted vector and the groundwater pressure-velocity time series sum weighted vector are calculated by position point.

6. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 5 is characterized in that: The groundwater temperature decoupling unit is used to perform feature decoupling on the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of groundwater temperature local spatial implicit correlation feature vectors as a sequence of groundwater temperature local spatial implicit correlation features.

7. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 6 is characterized in that: The groundwater multi-parameter interactive fusion unit comprises: A nonlinear transformation processing subunit, used for performing a nonlinear transformation on the groundwater pressure-flow velocity time series collaborative alignment fusion feature vector to obtain a groundwater pressure-flow velocity time series collaborative alignment fusion feature vector after nonlinear transformation; The association interaction processing subunit is used to calculate the association interaction vectors between the groundwater pressure-flow velocity time series coordinated alignment fusion feature vector after the nonlinear transformation and each groundwater temperature local space implicit association feature vector in the sequence of the groundwater temperature local space implicit association feature vector to obtain a sequence of groundwater spatiotemporal multi-parameter association interaction vectors; A groundwater spatiotemporal multi-parameter characteristic interaction energy factor calculation subunit, used to calculate the characteristic interaction energy factor of each groundwater spatiotemporal multi-parameter associated interaction vector in the sequence of groundwater spatiotemporal multi-parameter associated interaction vectors to obtain a sequence of groundwater spatiotemporal multi-parameter characteristic interaction energy factors; A normalization processing subunit, used for inputting the sequence of groundwater spatiotemporal multi-parameter characteristic interaction energy factors into a Sigmoid function to obtain a sequence of normalized groundwater spatiotemporal multi-parameter characteristic interaction energy factors; The groundwater spatiotemporal multi-parameter fusion processing subunit is used to use the sequence of normalized groundwater spatiotemporal multi-parameter characteristic interaction energy factors as a sequence of weights, calculate the position-weighted sum of the sequence of groundwater spatiotemporal multi-parameter associated interaction vectors to obtain a groundwater spatiotemporal multi-parameter fusion representation vector as the groundwater spatiotemporal multi-parameter fusion representation.

8. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 7 is characterized in that: The groundwater spatiotemporal multi-parameter characteristic interaction energy factor calculation subunit is used to: Respectively calculating the mean and variance of the groundwater spatiotemporal multi-parameter correlation interaction vector to obtain the groundwater spatiotemporal multi-parameter correlation interaction mean and the groundwater spatiotemporal multi-parameter correlation interaction variance; Calculating the positional difference between the groundwater spatiotemporal multi-parameter correlation interaction vector and the groundwater spatiotemporal multi-parameter correlation interaction mean to obtain a groundwater spatiotemporal multi-parameter correlation interaction difference vector; Calculating the fourth power of each eigenvalue in the groundwater spatiotemporal multi-parameter associated interactive difference vector to obtain a groundwater spatiotemporal multi-parameter associated interactive modulation difference vector; Calculating the expected value of the groundwater spatiotemporal multi-parameter correlation interaction modulation difference vector to obtain the groundwater spatiotemporal multi-parameter correlation interaction expected value; Dividing the groundwater spatiotemporal multi-parameter correlation interaction expected value by the square of the groundwater spatiotemporal multi-parameter correlation interaction variance to obtain the groundwater spatiotemporal multi-parameter correlation interaction kurtosis value; Calculating the sum of the groundwater spatiotemporal multi-parameter correlation interaction variance and the hyperparameter to obtain the groundwater spatiotemporal multi-parameter correlation interaction first energy factor; Calculate the square of the difference between the groundwater spatiotemporal multi-parameter correlation interaction kurtosis value and the groundwater spatiotemporal multi-parameter correlation interaction mean value to obtain the groundwater spatiotemporal multi-parameter correlation interaction difference value; The groundwater spatiotemporal multi-parameter correlation interaction difference value is added to obtain the groundwater spatiotemporal multi-parameter correlation interaction second energy factor; The first energy factor of the groundwater spatiotemporal multi-parameter correlation interaction is divided by the second energy factor of the groundwater spatiotemporal multi-parameter correlation interaction to obtain the groundwater spatiotemporal multi-parameter characteristic interaction energy factor corresponding to the groundwater spatiotemporal multi-parameter correlation interaction vector.

9. The high-precision monitoring system for groundwater flow velocity and direction based on multi-parameter fusion technology according to claim 8 is characterized in that: The flow direction monitoring result generating module is used to: input the groundwater spatiotemporal multi-parameter fusion representation vector into the decoder-based groundwater flow direction monitoring module to obtain the flow direction monitoring result, which is the short-term predicted decoding value of the groundwater flow direction.

10. A high-precision monitoring method for groundwater flow velocity and direction based on multi-parameter fusion technology, characterized in that: include: Obtaining a time series of groundwater pressure, a time series of groundwater velocity, and a depth series of groundwater temperature values; Extracting a groundwater pressure time series implicit correlation feature vector, a groundwater velocity time series implicit correlation feature vector and a groundwater temperature space implicit correlation feature vector from the groundwater pressure time series, the groundwater velocity time series and the groundwater temperature depth series respectively; Performing fine-grained optimal fusion interaction on the implicit correlation feature vector of the groundwater pressure time series and the implicit correlation feature vector of the groundwater velocity time series to obtain a groundwater pressure-velocity time series collaborative fusion feature; The groundwater pressure-velocity time series collaborative fusion feature and the groundwater temperature spatial implicit correlation feature vector are subjected to local and global collaborative fusion interaction to obtain a groundwater spatiotemporal multi-parameter fusion representation, wherein the groundwater pressure-velocity time series collaborative fusion feature and the groundwater temperature spatial implicit correlation feature vector are subjected to local and global collaborative fusion interaction, including: feature decoupling of the groundwater temperature spatial implicit correlation feature vector to obtain a sequence of groundwater temperature local spatial implicit correlation features; feature interactive fusion of the groundwater pressure-velocity time series collaborative fusion feature and the sequence of groundwater temperature local spatial implicit correlation features to obtain the groundwater spatiotemporal multi-parameter fusion representation; Based on the groundwater spatiotemporal multi-parameter fusion representation, the flow direction monitoring result is obtained.

Images