Water quality prediction method and system based on multi-scale feature fusion

Abstract

The invention discloses a water quality prediction method and system based on multi-scale feature fusion, and the method comprises the steps: carrying out the fusion of all multi-scale features in a multi-scale feature set according to a preset multi-scale feature interaction and fusion mechanism, and obtaining and generating a fused multi-scale feature; performing iterative training on the multi-scale water quality prediction sub-model according to the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features to obtain a target multi-scale water quality prediction sub-model; real-time fusion multi-scale features corresponding to the obtained real-time multi-source heterogeneous water quality data are input into a target multi-scale water quality prediction sub-model, the target multi-scale water quality prediction sub-model outputs corresponding sub-prediction results, and all sub-prediction modules are fused according to a preset dynamic weighted integration strategy. And obtaining a final water quality prediction result. The method effectively overcomes the defects of poor adaptability and limited precision of a traditional single time scale prediction model, and significantly improves the accuracy of water quality prediction.

Description

Technical Field

[0001] This invention belongs to the field of water quality prediction technology, and in particular relates to a water quality prediction method and system based on multi-scale feature fusion. Background Technology

[0002] Water quality forecasting is a key technology in water environment management and pollution control. Accurate prediction of multiple water body indicators provides crucial information for water resource allocation, pollution source tracing, and emergency decision-making. Traditional water quality forecasting methods often rely on single-time-scale statistical or mechanistic models, such as time series analysis, regression models, or simulation methods based on hydrology-water quality coupling. While these methods have practical value under certain conditions, they often struggle to effectively address the complex needs of fusing multi-source, heterogeneous, and multi-scale data in real-world environments, especially in cases of sudden water quality changes or multi-factor coupling, where forecast accuracy is limited.

[0003] In recent years, with the development of machine learning technology, algorithms such as Support Vector Machine (SVM), Random Forest, and Long Short-Term Memory (LSTM) networks have been gradually introduced into the field of water quality prediction, improving prediction performance to some extent. However, most of these methods are still limited to modeling at a single time scale and fail to fully explore the complementary information hidden in different time granularities (such as hourly, daily, and weekly). On the other hand, the heterogeneity and spatiotemporal inconsistency among multi-source data (such as water quality parameters, meteorological elements, and hydrological dynamics) also increase the difficulty of feature fusion and model generalization.

[0004] While some existing technologies have attempted to introduce multi-scale analysis concepts, such as decomposing time series using wavelet transform or constructing hierarchical prediction structures, the following main problems still exist: First, there is a lack of unified representation and interaction mechanisms at the feature level, resulting in insufficient fusion of multi-scale information; second, the model structure is often statically fixed, making it difficult to adapt to the differences and correlations between prediction tasks at different time scales; and third, the uncertainty and reliability of prediction results are not systematically considered, limiting its decision support capabilities in real-world environments.

[0005] Therefore, there is an urgent need for a water quality prediction method that can effectively integrate multi-scale features, adapt to prediction tasks of different time granularities, and has the ability to quantify uncertainty, so as to improve the accuracy and robustness of prediction models under complex water environment conditions. Summary of the Invention

[0006] This invention provides a water quality prediction method and system based on multi-scale feature fusion to solve the technical problem of inaccurate water quality prediction results in existing systems.

[0007] In a first aspect, the present invention provides a water quality prediction method based on multi-scale feature fusion, comprising: Acquire multi-source heterogeneous water quality data and preprocess the multi-source heterogeneous water quality data to obtain target multi-source heterogeneous water quality data; Based on the target multi-source heterogeneous water quality data, multi-scale water quality prediction sub-models corresponding to different time scales are constructed respectively, wherein the time scales include at least hourly, daily and weekly levels; Based on various time scales, feature mapping processing is performed on the target multi-source heterogeneous water quality data respectively, and the features of different time scales are transformed into a unified representation space to form a multi-scale feature set. The multi-scale features in the multi-scale feature set are fused according to the preset multi-scale feature interaction and fusion mechanism to obtain the generated fused multi-scale features; The multi-scale water quality prediction sub-model is iteratively trained based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features to obtain the target multi-scale water quality prediction sub-model. The real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data are input into the target multi-scale water quality prediction sub-model. The target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results and fuses each sub-prediction module according to a preset dynamic weighted integration strategy to obtain the final water quality prediction result.

[0008] Secondly, the present invention provides a water quality prediction system based on multi-scale feature fusion, comprising: The acquisition module is configured to acquire multi-source heterogeneous water quality data and preprocess the multi-source heterogeneous water quality data to obtain target multi-source heterogeneous water quality data. The construction module is configured to construct multi-scale water quality prediction sub-models corresponding to different time scales based on the target multi-source heterogeneous water quality data, wherein the time scales include at least hourly, daily, and weekly levels; The conversion module is configured to perform feature mapping processing on the target multi-source heterogeneous water quality data based on various time scales, and convert the features of different time scales to a unified representation space to form a multi-scale feature set. The fusion module is configured to fuse the various multi-scale features in the multi-scale feature set according to a preset multi-scale feature interaction and fusion mechanism to obtain a fused multi-scale feature. The training module is configured to iteratively train the multi-scale water quality prediction sub-model based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features, so as to obtain the target multi-scale water quality prediction sub-model. The output module is configured to input the real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data into the target multi-scale water quality prediction sub-model. The target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results and fuses each sub-prediction module according to a preset dynamic weighted integration strategy to obtain the final water quality prediction result.

[0009] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the water quality prediction method based on multi-scale feature fusion according to any embodiment of the present invention.

[0010] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the water quality prediction method based on multi-scale feature fusion according to any embodiment of the present invention.

[0011] This application presents a water quality prediction method and system based on multi-scale feature fusion. It acquires and preprocesses heterogeneous water quality data from multiple sources, constructs multi-scale water quality prediction sub-models corresponding to different time scales, then performs feature mapping on the data to form a multi-scale feature set. Based on a pre-defined multi-scale feature interaction and fusion mechanism, it generates fused multi-scale features, which are then used to iteratively train the multi-scale water quality prediction sub-models to obtain the target model. Finally, a dynamic weighted integration strategy is used to fuse the prediction results of each sub-model to achieve water quality prediction. This method effectively overcomes the shortcomings of traditional single-time-scale prediction models, such as poor adaptability and limited accuracy, significantly improving the accuracy and reliability of water quality prediction. Furthermore, the multi-scale feature fusion and dynamic integration mechanism enhance the model's ability to capture water quality change patterns across different time dimensions, demonstrating good generalization performance and practical value. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart of a water quality prediction method based on multi-scale feature fusion provided in an embodiment of the present invention; Figure 2 A structural block diagram of a water quality prediction system based on multi-scale feature fusion is provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] Please see Figure 1 The diagram shows a flowchart of a water quality prediction method based on multi-scale feature fusion according to this application.

[0016] like Figure 1 As shown, the water quality prediction method based on multi-scale feature fusion specifically includes the following steps: Step S101: Obtain multi-source heterogeneous water quality data and preprocess the multi-source heterogeneous water quality data to obtain target multi-source heterogeneous water quality data.

[0017] In this step, multi-source data, including water quality, meteorology, and hydrology, are collected. Time series interpolation and multivariate imputation are used to process missing values, and indicators of different dimensions are standardized to the same time scale to obtain the target multi-source heterogeneous data.

[0018] Step S102: Based on the target multi-source heterogeneous water quality data, construct multi-scale water quality prediction sub-models corresponding to different time scales, wherein the time scales include at least hourly, daily, and weekly levels.

[0019] In this step, models are selected based on different time scales to obtain the first water quality prediction sub-model corresponding to hourly prediction, the second water quality prediction sub-model corresponding to daily prediction, and the third water quality prediction sub-model corresponding to weekly prediction. The first water quality prediction sub-model, the second water quality prediction sub-model, and the third water quality prediction sub-model are scored according to a preset model scoring function, and a first target water quality prediction sub-model, a second target water quality prediction sub-model, and a third target water quality prediction sub-model are selected based on the scoring results. The expression for the model scoring function is as follows: , In the formula, The score for the i-th model is... Let be the historical accuracy of the i-th model. Let be the complexity of the i-th model. Let be the complexity of the i-th model. , , All are weighting coefficients, and ; A multi-scale architecture is designed for the first target water quality prediction sub-model, the second target water quality prediction sub-model, and the third target water quality prediction sub-model to obtain target multi-scale water quality prediction sub-models corresponding to different time scales. The expression for the multi-scale architecture of the target multi-scale water quality prediction sub-model is as follows: , In the formula, For the target multi-scale water quality prediction sub-model, The target is a multi-source heterogeneous water quality data sequence. For model weights, For model functions at different scales, For the number of models, For time scale.

[0020] It should be noted that for hourly short-term forecasts: the LSTM model is selected; for daily forecasts: the XGBoost model is selected; and for weekly forecasts: the Random Forest model is selected.

[0021] Step S103: Based on each time scale, feature mapping processing is performed on the target multi-source heterogeneous water quality data respectively, and the features of different time scales are transformed into a unified representation space to form a multi-scale feature set.

[0022] In this step, the expression for feature mapping processing of the target multi-source heterogeneous water quality data is as follows: , In the formula, For the mapped features, Original features For time scale, This is a mapping function.

[0023] Step S104: The multi-scale features in the multi-scale feature set are fused according to the preset multi-scale feature interaction and fusion mechanism to obtain the generated fused multi-scale features.

[0024] In this step, the expression for fusing multi-scale features is: , In the formula, For the mapped features, For time scale, For feature interaction sub-functions, The target is a multi-source heterogeneous water quality data sequence. To integrate multi-scale features.

[0025] Step S105: Iteratively train the multi-scale water quality prediction sub-model based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features to obtain the target multi-scale water quality prediction sub-model.

[0026] Step S106: Input the real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data into the target multi-scale water quality prediction sub-model. The target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results and fuses each sub-prediction module according to the preset dynamic weighted integration strategy to obtain the final water quality prediction result.

[0027] In this step, the expression for the dynamic weighted integration strategy is: , , In the formula, For the final water quality prediction results, For the sub-prediction results, The weights of the sub-prediction results, For frequency knowledge transfer, For semantic association features, This is a dynamic weighted correlation function. These are the model parameters from the previous time step. These are the model weights.

[0028] In summary, the method of this application acquires and preprocesses multi-source heterogeneous water quality data, constructs multi-scale water quality prediction sub-models corresponding to different time scales, then performs feature mapping processing on the data to form a multi-scale feature set, and generates fused multi-scale features based on a preset multi-scale feature interaction and fusion mechanism. Based on this, the multi-scale water quality prediction sub-model is iteratively trained to obtain the target model. Finally, the results of each sub-prediction are fused through a dynamic weighted integration strategy to achieve water quality prediction. This method effectively overcomes the shortcomings of traditional single-time-scale prediction models, such as poor adaptability and limited accuracy, significantly improving the accuracy and reliability of water quality prediction. Furthermore, the multi-scale feature fusion and dynamic integration mechanism enhance the model's ability to capture water quality change patterns across different time dimensions, demonstrating good generalization performance and practical value.

[0029] Please see Figure 2 The diagram shows a structural block diagram of a water quality prediction system based on multi-scale feature fusion according to this application.

[0030] like Figure 2As shown, the water quality prediction system 200 includes an acquisition module 210, a construction module 220, a conversion module 230, a fusion module 240, a training module 250, and an output module 260.

[0031] The system includes the following modules: an acquisition module 210, configured to acquire multi-source heterogeneous water quality data and preprocess the data to obtain target multi-source heterogeneous water quality data; a construction module 220, configured to construct multi-scale water quality prediction sub-models corresponding to different time scales based on the target multi-source heterogeneous water quality data, wherein the time scales include at least hourly, daily, and weekly levels; a conversion module 230, configured to perform feature mapping processing on the target multi-source heterogeneous water quality data at each time scale, converting the features at different time scales to a unified representation space to form a multi-scale feature set; and a fusion module 240, configured to perform multi-scale feature interaction and fusion according to a preset multi-scale feature interaction and fusion mechanism. The multi-scale features in the multi-scale feature set are fused to obtain fused multi-scale features; the training module 250 is configured to iteratively train the multi-scale water quality prediction sub-model based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features to obtain the target multi-scale water quality prediction sub-model; the output module 260 is configured to input the real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data into the target multi-scale water quality prediction sub-model, the target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results, and fuses the various sub-prediction modules according to a preset dynamic weighted integration strategy to obtain the final water quality prediction result.

[0032] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0033] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the water quality prediction method based on multi-scale feature fusion in any of the above method embodiments. In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows: Acquire multi-source heterogeneous water quality data and preprocess the multi-source heterogeneous water quality data to obtain target multi-source heterogeneous water quality data; Based on the target multi-source heterogeneous water quality data, multi-scale water quality prediction sub-models corresponding to different time scales are constructed respectively, wherein the time scales include at least hourly, daily and weekly levels; Based on various time scales, feature mapping processing is performed on the target multi-source heterogeneous water quality data respectively, and the features of different time scales are transformed into a unified representation space to form a multi-scale feature set. The multi-scale features in the multi-scale feature set are fused according to the preset multi-scale feature interaction and fusion mechanism to obtain the generated fused multi-scale features; The multi-scale water quality prediction sub-model is iteratively trained based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features to obtain the target multi-scale water quality prediction sub-model. The real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data are input into the target multi-scale water quality prediction sub-model. The target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results and fuses each sub-prediction module according to a preset dynamic weighted integration strategy to obtain the final water quality prediction result.

[0034] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of the water quality prediction system based on multi-scale feature fusion. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely disposed relative to a processor, which can be connected to the water quality prediction system based on multi-scale feature fusion via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0035] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3 As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby implementing the water quality prediction method based on multi-scale feature fusion as described in the above embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the water quality prediction system based on multi-scale feature fusion. The output device 340 may include a display screen or other display device.

[0036] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0037] In one implementation, the above-described electronic device is applied to a water quality prediction system based on multi-scale feature fusion, for a client, and includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire multi-source heterogeneous water quality data and preprocess the multi-source heterogeneous water quality data to obtain target multi-source heterogeneous water quality data; Based on the target multi-source heterogeneous water quality data, multi-scale water quality prediction sub-models corresponding to different time scales are constructed respectively, wherein the time scales include at least hourly, daily and weekly levels; Based on various time scales, feature mapping processing is performed on the target multi-source heterogeneous water quality data respectively, and the features of different time scales are transformed into a unified representation space to form a multi-scale feature set. The multi-scale features in the multi-scale feature set are fused according to the preset multi-scale feature interaction and fusion mechanism to obtain the generated fused multi-scale features; The multi-scale water quality prediction sub-model is iteratively trained based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features to obtain the target multi-scale water quality prediction sub-model. The real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data are input into the target multi-scale water quality prediction sub-model. The target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results and fuses each sub-prediction module according to a preset dynamic weighted integration strategy to obtain the final water quality prediction result.

[0038] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0039] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A water quality prediction method based on multi-scale feature fusion, characterized in that, include: Acquire multi-source heterogeneous water quality data and preprocess the multi-source heterogeneous water quality data to obtain target multi-source heterogeneous water quality data; Based on the target multi-source heterogeneous water quality data, multi-scale water quality prediction sub-models corresponding to different time scales are constructed respectively, wherein the time scales include at least hourly, daily and weekly levels; Based on various time scales, feature mapping processing is performed on the target multi-source heterogeneous water quality data respectively, and the features of different time scales are transformed into a unified representation space to form a multi-scale feature set. The multi-scale features in the multi-scale feature set are fused according to the preset multi-scale feature interaction and fusion mechanism to obtain the generated fused multi-scale features; The multi-scale water quality prediction sub-model is iteratively trained based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features to obtain the target multi-scale water quality prediction sub-model. The real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data are input into the target multi-scale water quality prediction sub-model. The target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results and fuses each sub-prediction module according to a preset dynamic weighted integration strategy to obtain the final water quality prediction result.

2. The water quality prediction method based on multi-scale feature fusion according to claim 1, characterized in that, The construction of water quality prediction sub-models corresponding to different time scales based on the target multi-source heterogeneous water quality data includes: Based on different time scales, the model is selected to obtain the first water quality prediction sub-model corresponding to hourly prediction, the second water quality prediction sub-model corresponding to daily prediction, and the third water quality prediction sub-model corresponding to weekly prediction. The first water quality prediction sub-model, the second water quality prediction sub-model, and the third water quality prediction sub-model are scored according to a preset model scoring function, and a first target water quality prediction sub-model, a second target water quality prediction sub-model, and a third target water quality prediction sub-model are selected based on the scoring results. The expression for the model scoring function is as follows: ， In the formula, The score for the i-th model is... Let be the historical accuracy of the i-th model. Let be the complexity of the i-th model. Let be the complexity of the i-th model. , , All are weighting coefficients, and ; A multi-scale architecture is designed for the first target water quality prediction sub-model, the second target water quality prediction sub-model, and the third target water quality prediction sub-model to obtain target multi-scale water quality prediction sub-models corresponding to different time scales. The expression for the multi-scale architecture of the target multi-scale water quality prediction sub-model is as follows: ， In the formula, For the target multi-scale water quality prediction sub-model, The target is a multi-source heterogeneous water quality data sequence. For model weights, For model functions at different scales, For the number of models, For time scale.

3. The water quality prediction method based on multi-scale feature fusion according to claim 1, characterized in that, in, The expressions for feature mapping processing of the target multi-source heterogeneous water quality data are as follows: ， In the formula, For the mapped features, Original features For time scale, This is a mapping function.

4. The water quality prediction method based on multi-scale feature fusion according to claim 1, characterized in that, The expression for fusing multi-scale features is: ， In the formula, For the mapped features, For time scale, For feature interaction sub-functions, The target is a multi-source heterogeneous water quality data sequence. To integrate multi-scale features.

5. The water quality prediction method based on multi-scale feature fusion according to claim 1, characterized in that, The expression for the dynamic weighted integration strategy is: ， ， In the formula, For the final water quality prediction results, For the sub-prediction results, The weights of the sub-prediction results, For frequency knowledge transfer, For semantic association features, It is a dynamic weighted correlation function. These are the model parameters from the previous time step. These are the model weights.

6. A water quality prediction system based on multi-scale feature fusion, characterized in that, include: The acquisition module is configured to acquire multi-source heterogeneous water quality data and preprocess the multi-source heterogeneous water quality data to obtain target multi-source heterogeneous water quality data. The construction module is configured to construct multi-scale water quality prediction sub-models corresponding to different time scales based on the target multi-source heterogeneous water quality data, wherein the time scales include at least hourly, daily, and weekly levels; The conversion module is configured to perform feature mapping processing on the target multi-source heterogeneous water quality data based on various time scales, and convert the features of different time scales to a unified representation space to form a multi-scale feature set. The fusion module is configured to fuse the various multi-scale features in the multi-scale feature set according to a preset multi-scale feature interaction and fusion mechanism to obtain a fused multi-scale feature. The training module is configured to iteratively train the multi-scale water quality prediction sub-model based on the fused multi-scale features and the water quality labels corresponding to the fused multi-scale features, so as to obtain the target multi-scale water quality prediction sub-model. The output module is configured to input the real-time fused multi-scale features corresponding to the acquired real-time multi-source heterogeneous water quality data into the target multi-scale water quality prediction sub-model. The target multi-scale water quality prediction sub-model outputs the corresponding sub-prediction results and fuses each sub-prediction module according to a preset dynamic weighted integration strategy to obtain the final water quality prediction result.

7. An electronic device, characterized in that, include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the water quality prediction method based on multi-scale feature fusion as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the water quality prediction method based on multi-scale feature fusion as described in any one of claims 1 to 5.

Images