Intelligent water quality prediction method and system based on multi-frequency adaptive adjustment

Abstract

The invention discloses an intelligent water quality prediction method and system based on multi-frequency adaptive adjustment, and the method comprises the steps: constructing a rainfall-season-spatial-temporal scale three-dimensional monitoring frequency matrix according to first target multi-source heterogeneous data by adopting a preset dynamic weight fusion strategy; constructing a confidence-error dual-drive frequency adaptive adjustment strategy according to the three-dimensional monitoring frequency matrix, and correcting a current acquisition frequency according to the frequency adaptive adjustment strategy to obtain a target acquisition frequency; and acquiring second target multi-source heterogeneous data of the water quality monitoring area under the target acquisition frequency, inputting the second target multi-source heterogeneous data into a preset water quality prediction model, and outputting by the water quality prediction model to obtain a prediction result corresponding to the second target multi-source heterogeneous data. The practical value of water quality monitoring and early warning is greatly improved, and powerful technical support is provided for water environment protection and drinking water safety.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of water quality prediction, and particularly relates to a water quality intelligent prediction method and system based on multi-frequency adaptive adjustment. BACKGROUND

[0002] Water quality prediction is one of the core technologies of water environment management and protection, and is of great significance for preventing water environment pollution risk, ensuring drinking water safety and maintaining water ecological system health. The traditional online water quality monitoring system adopts a fixed frequency data acquisition and reporting mode, which can obtain continuous monitoring data, but has significant limitations: on the one hand, during the stable period of water quality, fixed frequency acquisition may lead to data redundancy, causing waste of communication, storage and computing resources, and shortening the service life of the monitoring equipment; on the other hand, during the critical period of rainfall, pollution, etc. causing water quality to fluctuate dramatically, the fixed sampling frequency may not be able to capture sudden, short-duration pollution events, resulting in delayed or missed early warning. SUMMARY

[0003] The application provides a water quality intelligent prediction method and system based on multi-frequency adaptive adjustment, which is used to solve the technical problem that sudden, short-duration water pollution events cannot be captured, resulting in delayed or missed early warning.

[0004] In a first aspect, the application provides a water quality intelligent prediction method based on multi-frequency adaptive adjustment, comprising: Obtaining first multi-source heterogeneous data of a water quality monitoring area under a current acquisition frequency, and performing interpolation and multivariate interpolation processing on missing values of the first multi-source heterogeneous data using a time series, standardizing different dimension indicators to the same time scale to obtain first target multi-source heterogeneous data; According to the first target multi-source heterogeneous data, a pre-set dynamic weight fusion strategy is used to construct a three-dimensional monitoring frequency matrix of rainfall-season-spatial and temporal scale; According to the three-dimensional monitoring frequency matrix, a confidence-error double-driven frequency adaptive adjustment strategy is constructed, and the current acquisition frequency is corrected according to the frequency adaptive adjustment strategy to obtain a target acquisition frequency; Obtaining second target multi-source heterogeneous data of the water quality monitoring area under the target acquisition frequency, and inputting the second target multi-source heterogeneous data into a pre-set water quality prediction model, and the water quality prediction model outputs a prediction result corresponding to the second target multi-source heterogeneous data.

[0005] In a second aspect, the application provides a water quality intelligent prediction system based on multi-frequency adaptive adjustment, comprising: The acquisition module is configured to acquire first multi-source heterogeneous data of a water quality monitoring area at a current acquisition frequency, and perform interpolation and multivariate interpolation processing on missing values of the first multi-source heterogeneous data by using a time sequence, normalize different dimension index standards to the same time scale, and obtain first target multi-source heterogeneous data; The construction module is configured to construct a rainfall-season-spatiotemporal scale three-dimensional monitoring frequency matrix by using a preset dynamic weight fusion strategy according to the first target multi-source heterogeneous data; The correction module is configured to construct a confidence-error double-driven frequency adaptive adjustment strategy according to the three-dimensional monitoring frequency matrix, and correct the current acquisition frequency according to the frequency adaptive adjustment strategy to obtain a target acquisition frequency. The output module is configured to acquire second target multi-source heterogeneous data of the water quality monitoring area at the target acquisition frequency, and input the second target multi-source heterogeneous data into a preset water quality prediction model, and output a prediction result corresponding to the second target multi-source heterogeneous data by the water quality prediction model.

[0006] In a third aspect, an electronic device is provided, which includes at least one processor and a memory connected with the at least one processor in communication, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the steps of the water quality intelligent prediction method based on multi-frequency adaptive adjustment according to any one of the embodiments of the present application.

[0007] In a fourth aspect, the present application further provides a computer readable storage medium having a computer program stored thereon, and the program instructions are executed by a processor to enable the processor to perform the steps of the water quality intelligent prediction method based on multi-frequency adaptive adjustment according to any one of the embodiments of the present application.

[0008] The water quality intelligent prediction method and system based on multi-frequency adaptive adjustment according to the present application establish a complete closed-loop system from data acquisition to prediction output by constructing a three-dimensional monitoring frequency matrix and a double-driven frequency adjustment strategy, can dynamically adjust the data acquisition frequency according to real-time monitoring data and prediction result confidence, reduce the monitoring cost, effectively solve the problems of non-uniform spatiotemporal scale of multi-source heterogeneous data and difficulty in quantifying influence weight by introducing an attention mechanism and a dynamic weight fusion strategy, improve the feature representation efficiency, combine the prediction confidence with the historical error to construct a frequency adjustment quantization model, overcome the hysteresis of the traditional method which simply relies on the historical error, shorten the response time of the system in a sudden water pollution event, greatly improve the practical value of water quality monitoring and early warning, and provide strong technical support for water environment protection and drinking water safety. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without any creative effort.

[0010] Figure 1 A flow chart of a water quality intelligent prediction method based on multi-frequency adaptive adjustment provided by an embodiment of the present application is shown in Figure 2 A structural block diagram of a water quality intelligent prediction system based on multi-frequency adaptive adjustment provided by an embodiment of the present application is shown in Figure 3 A structural schematic diagram of an electronic device provided by an embodiment of the present application is shown in DETAILED DESCRIPTION

[0011] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to make a clear and complete description of the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort are within the protection scope of the present application.

[0012] Please refer to Figure 1 , which shows a flow chart of a water quality intelligent prediction method based on multi-frequency adaptive adjustment.

[0013] As shown in Figure 1 , the water quality intelligent prediction method based on multi-frequency adaptive adjustment specifically includes the following steps: Step S101, acquiring first multi-source heterogeneous data of a water quality monitoring area under a current acquisition frequency, and adopting time series to interpolate and multi-variable interpolation to handle missing values of the first multi-source heterogeneous data, normalizing different dimension index to the same time scale to obtain first target multi-source heterogeneous data.

[0014] In this step, multi-source data including water quality, weather, hydrology, etc. are collected, time series interpolation and multi-variable interpolation are adopted to handle missing values, and different dimension indexes are normalized to the same time scale to obtain the first target multi-source heterogeneous data.

[0015] Step S102, according to the first target multi-source heterogeneous data, adopting a preset dynamic weight fusion strategy to construct a three-dimensional monitoring frequency matrix of rainfall-season-spatial and temporal scale.

[0016] In this step, based on the rainfall data in the first target multi-source heterogeneous data, the current rainfall intensity is calculated, and according to the preset rainfall intensity grading threshold, the rainfall intensity grade is determined, which includes light rainfall, moderate rainfall and heavy rainfall; Based on the time data in the first target multi-source heterogeneous data, the current season type is determined, which includes spring snowmelt period, summer rainstorm period, autumn stable period and winter low temperature period, and the preset season weight factor is configured for each season type; Based on the predicted demand and environmental complexity, the target time scale is determined, which includes hour level, day level, week level and month level; Based on the attention mechanism, the influence weight of the meteorological mutation factor and the human activity factor is dynamically calculated, wherein the meteorological mutation factor includes short-time rainfall intensity change rate and temperature sudden change index, and the human activity factor includes holiday identification and industrial drainage time mode; Based on the rainfall intensity grade, season type and time scale, an initial three-dimensional matrix is constructed, and according to the initial three-dimensional matrix, the human activity factor and the meteorological mutation factor, the three-dimensional monitoring frequency matrix is constructed, wherein the expression of the initial three-dimensional matrix is: , In the formula, is a three-dimensional monitoring frequency matrix, is an initial three-dimensional matrix constructed based on rainfall intensity grade, season type and time scale, is a meteorological weight matrix, is a human activity weight matrix, is an attention weighting function, is a meteorological mutation factor, is a human activity factor, are all dynamic fusion coefficients, and .

[0017] Step S103, according to the three-dimensional monitoring frequency matrix, a confidence-error double-driven frequency adaptive adjustment strategy is constructed, and the current collection frequency is corrected according to the frequency adaptive adjustment strategy to obtain the target collection frequency.

[0018] In this step, based on the historical prediction data, the mean absolute error and the root mean square error of the water quality prediction model are calculated, and the prediction error is smoothed by using the exponential weighted moving average algorithm to obtain the error index ; The Monte Carlo dropout method is used to statistically analyze the multiple prediction results of the water quality prediction model, and the confidence level of the prediction result is calculated; According to the error index and a confidence level calculating a dynamic frequency adjustment amplitude; correcting a current acquisition frequency according to the dynamic frequency adjustment amplitude to obtain a target acquisition frequency, wherein an expression for calculating the target acquisition frequency is: , In the formula, is the dynamic frequency adjustment amplitude, is the current acquisition frequency, is the target acquisition frequency.

[0019] It should be noted that the expression for calculating the dynamic frequency adjustment amplitude is: , In the formula, is the matrix element value corresponding to the current rainfall level i, the seasonal type j, and the time scale k in the three-dimensional monitoring frequency matrix, is a basic adjustment coefficient, is a confidence weight factor.

[0020] Step S104, acquiring second target multi-source heterogeneous data of the water quality monitoring area under the target acquisition frequency, and inputting the second target multi-source heterogeneous data into a preset water quality prediction model, wherein the water quality prediction model outputs a prediction result corresponding to the second target multi-source heterogeneous data.

[0021] In summary, the method of the present application establishes a complete closed-loop system from data acquisition to prediction output by constructing a three-dimensional monitoring frequency matrix and a double-drive frequency adjustment strategy, can dynamically adjust the data acquisition frequency according to the real-time monitoring data and the prediction result confidence, reduces the monitoring cost, effectively solves the problem of non-uniform time and space scale of multi-source heterogeneous data and difficulty in quantifying influence weight by introducing the attention mechanism and the dynamic weight fusion strategy, improves the feature representation efficiency, combines the prediction confidence with the historical error to construct a frequency adjustment quantization model, overcomes the hysteresis of the traditional method which simply relies on the historical error, shortens the response time of the system in the sudden water pollution event, greatly improves the practical value of water quality monitoring and early warning, and provides strong technical support for water environment protection and drinking water safety.

[0022] Please refer to Figure 2 which shows a structure block diagram of a water quality intelligent prediction system based on multi-frequency adaptive adjustment.

[0023] As shown in Figure 2 , the water quality intelligent prediction system 200 includes an acquisition module 210, a construction module 220, a correction module 230, and an output module 240.

[0024] The acquisition module 210 is configured to acquire first multi-source heterogeneous data of a water quality monitoring area under a current acquisition frequency, and perform interpolation and multivariate interpolation processing on missing values of the first multi-source heterogeneous data by using a time sequence, normalize different dimension indicators to the same time scale, and obtain first target multi-source heterogeneous data. The construction module 220 is configured to construct a rainfall-season-spatiotemporal scale three-dimensional monitoring frequency matrix by using a preset dynamic weight fusion strategy according to the first target multi-source heterogeneous data. The correction module 230 is configured to construct a confidence-error double-driven frequency adaptive adjustment strategy according to the three-dimensional monitoring frequency matrix, and correct the current acquisition frequency according to the frequency adaptive adjustment strategy to obtain a target acquisition frequency. The output module 240 is configured to acquire second target multi-source heterogeneous data of the water quality monitoring area under the target acquisition frequency, and input the second target multi-source heterogeneous data into a preset water quality prediction model. The water quality prediction model outputs a prediction result corresponding to the second target multi-source heterogeneous data.

[0025] It should be understood that Figure 2 The modules described in the above Figure 1 The modules described in the above Figure 2 The modules described in the above

[0026] In some other embodiments, the present application also provides a computer readable storage medium having a computer program stored thereon, wherein the program instructions are executed by a processor to make the processor execute the water quality intelligent prediction method based on multi-frequency adaptive adjustment in any of the above method embodiments. As an implementation manner, the computer readable storage medium of the present application stores computer executable instructions, and the computer executable instructions are configured to: acquire first multi-source heterogeneous data of a water quality monitoring area under a current acquisition frequency, and perform interpolation and multivariate interpolation processing on missing values of the first multi-source heterogeneous data by using a time sequence, normalize different dimension indicators to the same time scale, and obtain first target multi-source heterogeneous data; construct a rainfall-season-spatiotemporal scale three-dimensional monitoring frequency matrix by using a preset dynamic weight fusion strategy according to the first target multi-source heterogeneous data; construct a confidence-error double-driven frequency adaptive adjustment strategy according to the three-dimensional monitoring frequency matrix, and correct the current acquisition frequency according to the frequency adaptive adjustment strategy to obtain a target acquisition frequency; Acquire second target multi-source heterogeneous data of the water quality monitoring area at the target acquisition frequency, and input the second target multi-source heterogeneous data into a preset water quality prediction model. The water quality prediction model outputs a prediction result corresponding to the second target multi-source heterogeneous data.

[0027] 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 multi-frequency adaptive water quality intelligent prediction system, etc. 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 configured relative to a processor, which can be connected to the multi-frequency adaptive water quality intelligent prediction system 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.

[0028] 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 3 Taking 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 realizing the water quality intelligent prediction method based on multi-frequency adaptive adjustment described in the above method 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 intelligent prediction system based on multi-frequency adaptive adjustment. The output device 340 may include a display screen or other display device.

[0029] 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.

[0030] As an implementation form, the electronic device is applied to a water quality intelligent prediction system based on multi-frequency adaptive adjustment, and is used for a client, comprising: at least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: obtain first multi-source heterogeneous data of a water quality monitoring area at a current collection frequency, and perform interpolation and multivariate interpolation processing on missing values of the first multi-source heterogeneous data by using a time sequence, so as to normalize different dimension index standards to the same time scale and obtain first target multi-source heterogeneous data; according to the first target multi-source heterogeneous data, a preset dynamic weight fusion strategy is used to construct a three-dimensional monitoring frequency matrix of rainfall-season-spatial and temporal scale; a confidence-error double-driven frequency adaptive adjustment strategy is constructed according to the three-dimensional monitoring frequency matrix, and the current collection frequency is corrected according to the frequency adaptive adjustment strategy to obtain a target collection frequency; obtain second target multi-source heterogeneous data of the water quality monitoring area at the target collection frequency, and input the second target multi-source heterogeneous data into a preset water quality prediction model, and the water quality prediction model outputs a prediction result corresponding to the second target multi-source heterogeneous data.

[0031] Through the description of the above implementation forms, those skilled in the art can clearly understand that each implementation form can be realized by means of software and necessary general hardware platform, and of course, it can also be realized by hardware. Based on such understanding, the above technical solutions essentially or say the part of the prior art that makes a contribution can be embodied in the form of a software product, which can be stored in a computer readable storage medium such as ROM / RAM, magnetic disk, optical disk, etc., and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the method of each embodiment or some part of the embodiment.

[0032] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for some technical features thereof; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A water quality intelligent prediction method based on multi-frequency adaptive adjustment, characterized in that, The method comprises the following steps: acquiring first multi-source heterogeneous data of a water quality monitoring area under a current acquisition frequency, and performing interpolation and multivariate interpolation processing on missing values of the first multi-source heterogeneous data by using a time sequence, normalizing different dimension index to the same time scale, and obtaining first target multi-source heterogeneous data; according to the first target multi-source heterogeneous data, a preset dynamic weight fusion strategy is adopted to construct a three-dimensional monitoring frequency matrix of rainfall-season-spatiotemporal scale; according to the three-dimensional monitoring frequency matrix, a confidence-error double-driven frequency adaptive adjustment strategy is constructed, and the current acquisition frequency is corrected according to the frequency adaptive adjustment strategy to obtain a target acquisition frequency; acquiring second target multi-source heterogeneous data of the water quality monitoring area under the target acquisition frequency, and inputting the second target multi-source heterogeneous data into a preset water quality prediction model, and the water quality prediction model outputs a prediction result corresponding to the second target multi-source heterogeneous data. 2.The water quality intelligent prediction method based on multi-frequency adaptive adjustment according to claim 1, characterized in that, The method comprises the following steps: based on rainfall data in the first target multi-source heterogeneous data, the current rainfall intensity is calculated, and the rainfall intensity grade is determined according to the preset rainfall intensity grading threshold, wherein the rainfall intensity grade includes light rainfall, moderate rainfall and heavy rainfall; based on the time data in the first target multi-source heterogeneous data, the current season type is determined, and the season type includes spring snowmelt period, summer rainstorm period, autumn stable period and winter low temperature period, and a preset season weight factor is configured for each season type; based on the prediction demand and the environmental complexity, a target time scale is determined, and the time scale includes hour level, day level, week level and month level; based on the attention mechanism, the influence weight of the meteorological mutation factor and the human activity factor is dynamically calculated, wherein the meteorological mutation factor includes short-time rainfall intensity change rate and temperature sudden change index, and the human activity factor includes holiday identification and industrial drainage time mode; an initial three-dimensional matrix is constructed based on the rainfall intensity grade, the season type and the time scale, and the three-dimensional monitoring frequency matrix is constructed based on the initial three-dimensional matrix, the human activity factor and the meteorological mutation factor, wherein the expression of the initial three-dimensional matrix is: ， wherein, is a three-dimensional monitoring frequency matrix, is an initial three-dimensional matrix constructed based on rainfall intensity levels, seasonal types and time scales, is a meteorological weight matrix, is a human activity weight matrix, is an attention weighting function, is a meteorological mutation factor, is a human activity factor, are dynamic fusion coefficients, and . 3.The water quality intelligent prediction method based on multi-frequency adaptive adjustment according to claim 1, characterized in that, The method comprises the following steps: Based on historical prediction data, the mean absolute error and root mean square error of the water quality prediction model are calculated, and the prediction error is smoothed by using an exponential weighted moving average algorithm to obtain an error index ; Adopt Monte Carlo dropout method, carry out statistics to multiple prediction results of the water quality prediction model, calculate the confidence level of the prediction result ; According to the error index and the confidence level Calculate the dynamic frequency adjustment range; according to the three-dimensional monitoring frequency matrix, a confidence-error double-driven frequency adaptive adjustment strategy is constructed, and the current acquisition frequency is corrected according to the frequency adaptive adjustment strategy to obtain a target acquisition frequency; ， In the formula, is the dynamic frequency adjustment amplitude, is the current acquisition frequency, is the target acquisition frequency.

4. The water quality intelligent prediction method based on multi-frequency adaptive adjustment according to claim 3, characterized in that, the current acquisition frequency is corrected according to the dynamic frequency adjustment amplitude to obtain a target acquisition frequency, wherein the expression for calculating the target acquisition frequency is: wherein, ， wherein is the matrix element value in the three-dimensional monitoring frequency matrix corresponding to the current rainfall intensity i, season type j and time scale k, is the base adjustment coefficient, is the confidence weight factor.

5. A multi-frequency adaptive adjustment-based water quality intelligent prediction system, characterized in that, the expression for calculating the dynamic frequency adjustment amplitude is: The method comprises the following steps: an acquisition module is configured to acquire first multi-source heterogeneous data of a water quality monitoring area under a current acquisition frequency, and perform interpolation and multivariate interpolation processing on missing values of the first multi-source heterogeneous data by using a time sequence, normalize different dimension index to the same time scale, and obtain first target multi-source heterogeneous data; The construction module is configured to construct a three-dimensional monitoring frequency matrix of rainfall-season-spatiotemporal scale according to the first target multi-source heterogeneous data and by using a preset dynamic weight fusion strategy. The correction module is configured to construct a confidence-error double-driven frequency self-adaptive adjustment strategy according to the three-dimensional monitoring frequency matrix, and correct the current collection frequency according to the frequency self-adaptive adjustment strategy to obtain a target collection frequency. The output module is configured to acquire second target multi-source heterogeneous data of a water quality monitoring area under the target collection frequency, and input the second target multi-source heterogeneous data into a preset water quality prediction model, and the water quality prediction model outputs a prediction result corresponding to the second target multi-source heterogeneous data.

6. An electronic device, comprising: Comprise: At least one processor, and a memory connected with the at least one processor in communication, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 4.

7. A computer readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the method of any one of claims 1 to 4.

Images