A smart water affairs and water conservancy algorithm development system and method based on artificial intelligence

Abstract

The present invention relates to the field of smart water affairs and water conservancy technology, and specifically discloses a smart water affairs and water conservancy algorithm development system and method based on artificial intelligence. The system collects water affairs data within a monitoring period and stores them in a database, calculates the warning ratio of water affairs data to warning values, and then obtains a water affairs analysis coefficient, and generates a water affairs analysis signal. The sliding window method is used to analyze the continuity of the signal, and the continuous moment ratio and the comprehensive mean ratio are calculated to obtain a water affairs warning coefficient, generate an AI analysis signal, obtain historical water affairs and meteorological data, and construct a smart water conservancy model with a recurrent neural network algorithm to predict future water affairs warning coefficients and generate a prediction warning signal. The present invention integrates multi-source data, uses an artificial intelligence algorithm for in-depth analysis, and effectively improves the intelligence level of water affairs and water conservancy management.

Description

Technical Field

[0001] The present invention relates to the field of smart water affairs and water conservancy technology, and specifically to a smart water affairs and water conservancy algorithm development system and method based on artificial intelligence. Background Art

[0002] With the acceleration of urbanization, the continuous growth of population and the impact of climate change, the water and water conservancy sectors are facing many severe challenges.

[0003] In the existing technology, some existing data analysis technologies can only perform simple statistical analysis, and are insufficient in mining the complex nonlinear relationships between data. They cannot accurately capture the complex interactions between meteorological factors, water conservancy project operations and other factors and water level and flow changes. Traditional early warning methods often simply set a few warning levels, which cannot accurately reflect the actual risk level faced by the water system. The existing water management system lacks the ability to self-learn and optimize, and cannot automatically improve algorithms and models based on new data and actual operating conditions.

[0004] By integrating multi-source data such as water level, flow, meteorology, and geography, the multi-source data is pre-processed to provide an accurate and reliable data basis for subsequent analysis. The neural network algorithm is used to conduct in-depth analysis of the integrated data, explore the complex relationships between the data, and discover the connection between water data and meteorological data. Through a multi-layer early warning classification strategy, the warning level is further subdivided according to the size of the water analysis coefficient and the early warning coefficient, which enables relevant departments to more accurately understand the degree of risk faced by the water system.

[0005] To this end, the present invention provides an artificial intelligence-based smart water affairs and water conservancy algorithm development system and method. Summary of the invention

[0006] The purpose of the present invention is to provide an artificial intelligence-based smart water affairs and water conservancy algorithm development system and method to solve the problems in the above background.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A method for developing smart water services and water conservancy algorithms based on artificial intelligence, comprising:

[0009] Collect water service data and store it in the database;

[0010] Obtain the water level value and flow value in the database at each monitoring moment during the monitoring period, calculate the water level value and flow value with the corresponding water level warning value and flow warning value respectively, and obtain the water level warning ratio Sw and flow warning ratio Ll, and further calculate the water affairs analysis coefficient Fx. If the water affairs analysis coefficient Fx ≥ the water affairs analysis threshold Fxz, generate a water affairs analysis signal;

[0011] Based on the water affairs analysis signal, the sliding window method is used to analyze the continuity of the water affairs analysis signal, obtain continuous monitoring moments, calculate the continuous moment ratio Lx, obtain the difference between the continuously monitored water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz, calculate the comprehensive mean ratio Zh, and calculate the water affairs early warning coefficient Yj based on the continuous moment ratio Lx and the comprehensive mean ratio Zh. The water affairs early warning coefficient Yj < the water affairs early warning coefficient threshold Yjz, and generate an AI analysis signal;

[0012] Based on AI analysis signals, historical water affairs data and meteorological data are obtained, a smart water conservancy model is constructed, and the model is used to predict and calculate the future water affairs warning coefficient Yj to obtain the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it indicates that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated.

[0013] As a further technical solution of the present invention: the water service analysis coefficient Fx is obtained as follows:

[0014] Based on the water level warning ratio Sw and flow warning ratio Ll, the water affairs analysis coefficient Fx is calculated through a weighted formula;

[0015] The water service analysis coefficient Fx is compared with the water service analysis threshold Fxz to generate a water service analysis signal.

[0016] As a further technical solution of the present invention: the water level warning ratio Sw is obtained as follows:

[0017] Obtain the water level value in the database at each monitoring moment during the monitoring period, make a difference between the water level value and the water level warning value, and take the absolute value to obtain the water level warning difference;

[0018] The water level warning difference is ratioed with the water level warning value to obtain the water level warning ratio, which is marked as Sw.

[0019] As a further technical solution of the present invention: the flow alarm ratio L1 is obtained as follows:

[0020] Obtain the flow value in the database at each monitoring moment during the monitoring period, make a difference between the flow value and the flow warning value, and take the absolute value to obtain the flow warning difference;

[0021] The flow warning difference is ratioed with the flow warning value to obtain the flow warning ratio, which is marked as L1.

[0022] As a further technical solution of the present invention: the water affairs analysis signal is generated in the following manner: if the water affairs analysis coefficient Fx≥water affairs analysis threshold Fxz, a water affairs analysis signal is generated.

[0023] As a further technical solution of the present invention: the water affairs early warning coefficient Yj is obtained as follows:

[0024] Based on the continuous time ratio Lx and the comprehensive mean ratio Zh, the water affairs early warning coefficient Yj is calculated;

[0025] By formula: , the water affairs early warning coefficient Yj is calculated, where ln(b1*Lx+b2*Zh+1.011) is a logarithmic function with base e, and b1 and b2 are preset proportional coefficients;

[0026] As a further technical solution of the present invention: the continuous time ratio Lx is obtained in the following manner:

[0027] Get the size of the sliding window, that is, the number of continuous moments, each time the sliding window is paused, and sum up all the continuous moment times to obtain the continuous moment sum;

[0028] Obtain the number of all monitoring moments within the monitoring period, perform ratio processing on the sum of the continuous moments and the number of all monitoring moments to obtain the continuous moment ratio, and mark the continuous moment ratio as Lx;

[0029] The method for obtaining the comprehensive mean ratio Zh is:

[0030] Obtain the difference between the water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz at each moment in the sliding window each time the sliding window is paused, and obtain the analysis coefficient difference;

[0031] The analysis coefficient differences when the sliding window is paused each time are summed and averaged to obtain the coefficient difference mean;

[0032] During the monitoring period, all coefficient difference means are summed and averaged to obtain the comprehensive coefficient difference mean;

[0033] The comprehensive coefficient difference mean is ratioed with the water affairs analysis coefficient threshold Fxz to obtain the comprehensive mean ratio, which is marked as Zh.

[0034] As a further technical solution of the present invention: the steps of analyzing the continuity of the water service analysis signal are as follows:

[0035] S1. Determine two consecutive monitoring moments as a sliding window, and place the sliding window at the starting position of the monitoring cycle, that is, the first monitoring moment;

[0036] S2, obtaining a water service analysis signal within the sliding window. If a water service analysis signal is generated at each monitoring moment within the sliding window, it is considered that the water service analysis signal is continuous within the sliding window;

[0037] S3, move the sliding window forward by one monitoring time, repeat S2, and judge the continuity of the water service analysis signal at the remaining monitoring time. If the sliding window moves to the next monitoring time and no water service analysis signal is generated, suspend monitoring;

[0038] S4. Obtain the monitoring time and water service analysis coefficient of the sliding window of suspended monitoring, continue to move the sliding window forward by one monitoring time, and repeat S2 and S3 until the sliding window covers the entire monitoring cycle.

[0039] As a further technical solution of the present invention: the construction method of the smart water conservancy model is as follows:

[0040] SS1, based on the historical water affairs data and meteorological data obtained, the model data set is integrated, and the data set is preprocessed to remove outliers and missing values;

[0041] SS2, using the recurrent neural network (RNN) architecture and ReLU as the activation function, the preprocessed data set is divided into a test set and a training set;

[0042] SS3, use the training set to train the model, and use the validation set to evaluate the model training results;

[0043] The prediction warning signal is generated as follows:

[0044] Based on the constructed smart water conservancy model, water conservancy data is predicted and analyzed according to future meteorological data;

[0045] Input future monitoring cycle data and future meteorological data, use the smart water conservancy model to calculate the future water conservancy early warning coefficient Yj, and obtain the water conservancy prediction coefficient Yc;

[0046] If the water affairs prediction coefficient Yc ≥ the water affairs early warning coefficient threshold Yjz, it means that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated.

[0047] An artificial intelligence-based smart water affairs and water conservancy algorithm development system, including:

[0048] Data acquisition module: collects water affairs data and stores it in the database;

[0049] Data analysis module: obtain the water level value and flow value in the database at each monitoring moment in the monitoring period, calculate the water level value and flow value with the corresponding water level warning value and flow warning value respectively, obtain the water level warning ratio Sw and flow warning ratio Ll, further calculate the water affairs analysis coefficient Fx, if the water affairs analysis coefficient Fx ≥ the water affairs analysis threshold Fxz, generate a water affairs analysis signal;

[0050] Early warning judgment module: Based on the water affairs analysis signal, the sliding window method is used to analyze the continuity of the water affairs analysis signal, obtain the continuous monitoring time, calculate the continuous time ratio Lx, obtain the difference between the continuously monitored water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz, calculate the comprehensive mean ratio Zh, and calculate the water affairs early warning coefficient Yj based on the continuous time ratio Lx and the comprehensive mean ratio Zh. The water affairs early warning coefficient Yj < the water affairs early warning coefficient threshold Yjz, and generate an AI analysis signal;

[0051] AI prediction module: Based on AI analysis signals, historical water affairs data and meteorological data are obtained to build a smart water conservancy model. The model is used to predict and calculate the future water affairs warning coefficient Yj to obtain the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it indicates that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated.

[0052] Beneficial effects of the present invention:

[0053] Collect water data within the monitoring period to ensure that the data sources are extensive and real-time. Store water level and flow values ​​in the database to facilitate subsequent centralized management and analysis. By calculating the water level warning ratio and flow warning ratio, and further deriving the water analysis coefficient, accurate quantitative analysis of water data is achieved. When the water analysis coefficient exceeds the threshold, a water analysis signal is generated.

[0054] Based on historical water affairs data and meteorological data, a smart water conservancy model is constructed, and a recurrent neural network algorithm is used to enable it to learn the long-term change laws and dynamic characteristics of the water affairs system. It can also predict future water affairs and calculate the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it means that the water affairs prediction coefficient Yc exceeds expectations, and a prediction warning signal is generated. Through timely warning, the probability and impact of water affairs disasters can be effectively reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The present invention will be further described below in conjunction with the accompanying drawings.

[0056] Figure 1 It is a flow chart of a method for developing smart water affairs and water conservancy algorithms based on artificial intelligence in the present invention;

[0057] Figure 2It is a module diagram of a smart water and water conservancy algorithm development system based on artificial intelligence. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] Example 1

[0060] See also Figure 1 As shown, the present invention is a method for developing smart water affairs and water conservancy algorithms based on artificial intelligence, comprising:

[0061] Step 1: Collect water affairs data and store the water affairs data in a database;

[0062] Use water level sensors and flow sensors to collect water service data within the monitoring period;

[0063] It should be noted that water affairs data include: water level value, flow value;

[0064] The water level value and flow value are stored in the database, wherein there are multiple water level values ​​and flow values ​​within the monitoring period;

[0065] Step 2: Obtain the water level value and flow value in the database at each monitoring moment in the monitoring period, calculate the water level value and flow value with the corresponding water level warning value and flow warning value respectively, and obtain the water level warning ratio Sw and flow warning ratio Ll, and further calculate the water affairs analysis coefficient Fx. If the water affairs analysis coefficient Fx ≥ the water affairs analysis threshold Fxz, generate a water affairs analysis signal;

[0066] Obtain the water level value in the database at each monitoring moment during the monitoring period, make the difference between the water level value and the water level warning value, and take the absolute value to obtain the water level warning difference;

[0067] The water level warning difference is processed as a ratio with the water level warning value to obtain the water level warning ratio, and the water level warning ratio is marked as Sw;

[0068] Obtain the flow value in the database at each monitoring moment during the monitoring period, make a difference between the flow value and the flow warning value, and take the absolute value to obtain the flow warning difference;

[0069] The flow warning difference is processed as a ratio with the flow warning value to obtain the flow warning ratio, and the flow warning ratio is marked as L1;

[0070] It should be noted that the water level warning ratio Sw and flow warning ratio Ll are calculated from the single data of each period within the monitoring period;

[0071] Calculate the water affairs analysis coefficient Fx based on the water level warning ratio Sw and the flow warning ratio Ll;

[0072] By formula: , calculate the water service analysis coefficient Fx, where a1 and a2 are preset proportional coefficients;

[0073] Compare the water service analysis coefficient Fx with the water service analysis threshold Fxz;

[0074] If the water affairs analysis coefficient Fx ≥ the water affairs analysis threshold Fxz, it indicates that the current water affairs analysis coefficient exceeds the target threshold, and a water affairs analysis signal is generated;

[0075] If the water affairs analysis coefficient Fx is less than the water affairs analysis threshold Fxz, it indicates that the current water affairs analysis coefficient is within the expected range and the change of the water affairs analysis coefficient still needs to be continuously monitored;

[0076] It should be noted that the water service analysis coefficient Fx is calculated from a single water level value and flow value. There are multiple water service analysis signals in the monitoring period. After the water service analysis signal is generated, the change of the water service analysis coefficient Fx calculated from the remaining water level value and flow value in the monitoring period is analyzed;

[0077] The technical solution of this embodiment is: collect water service data, store the water service data in a database, obtain the water level value and flow value of the water level data in the database for each time period within the monitoring period, calculate the water level value and flow value with the corresponding water level warning value and flow warning value respectively, and obtain the water level warning ratio Sw and flow warning ratio Ll, and further calculate the water service analysis coefficient Fx. If the water service analysis coefficient Fx≥water service analysis threshold Fxz, generate a water service analysis signal.

[0078] Example 2

[0079] Based on Example 1, the present invention is a method for developing smart water affairs and water conservancy algorithms based on artificial intelligence, and further includes:

[0080] Step 3: Based on the water affairs analysis signal, use the sliding window method to analyze the continuity of the water affairs analysis signal, obtain continuous monitoring moments, calculate the continuous moment ratio Lx, obtain the difference between the continuously monitored water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz, calculate the comprehensive mean ratio Zh, calculate the water affairs early warning coefficient Yj based on the continuous moment ratio Lx and the comprehensive mean ratio Zh, the water affairs early warning coefficient Yj<water affairs early warning coefficient threshold Yjz, and generate an AI analysis signal;

[0081] The sliding window method is used to analyze the continuity of water service analysis signals;

[0082] The steps to analyze the continuity of water service analysis signals using the sliding window method are as follows:

[0083] S1. Determine two consecutive monitoring moments as a sliding window, and place the sliding window at the starting position of the monitoring cycle, that is, the first monitoring moment;

[0084] S2, obtaining a water service analysis signal within the sliding window. If a water service analysis signal is generated at each monitoring moment within the sliding window, it is considered that the water service analysis signal is continuous within the sliding window;

[0085] S3, move the sliding window forward by one monitoring time, repeat S2, and judge the continuity of the water service analysis signal at the remaining monitoring time. If the sliding window moves to the next monitoring time and no water service analysis signal is generated, suspend monitoring;

[0086] S4, obtaining the monitoring time and water service analysis coefficient of the sliding window of suspended monitoring, continuing to move the sliding window forward by one monitoring time, and repeating S2 and S3 until the sliding window covers the entire monitoring cycle;

[0087] Get the size of the sliding window, that is, the number of continuous moments, each time the sliding window is paused, and sum up all the continuous moment times to obtain the continuous moment sum;

[0088] It should be noted that the total number of continuous moments refers to the number of continuous moments of the entire analysis process after the process of analyzing the continuity of the water service analysis signal by the sliding window method is completed;

[0089] Obtain the number of all monitoring moments within the monitoring period, perform ratio processing on the sum of the continuous moments and the number of all monitoring moments to obtain the continuous moment ratio, and mark the continuous moment ratio as Lx;

[0090] Obtain the difference between the water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz at each moment in the sliding window each time the sliding window is paused, and obtain the analysis coefficient difference;

[0091] The analysis coefficient differences when the sliding window is paused each time are summed and averaged to obtain the coefficient difference mean;

[0092] During the monitoring period, all coefficient difference means are summed and averaged to obtain the comprehensive coefficient difference mean;

[0093] The comprehensive coefficient difference mean is processed by ratio processing with the water affairs analysis coefficient threshold Fxz to obtain the comprehensive mean ratio, which is marked as Zh;

[0094] Based on the continuous time ratio Lx and the comprehensive mean ratio Zh, the water affairs early warning coefficient Yj is calculated;

[0095] By formula: , the water affairs early warning coefficient Yj is calculated, where ln(b1*Lx+b2*Zh+1.011) is a logarithmic function with base e, and b1 and b2 are preset proportional coefficients;

[0096] Compare the water affairs early warning coefficient Yj with the water affairs early warning coefficient threshold value Yjz;

[0097] If the water affairs warning coefficient Yj ≥ the water affairs warning coefficient threshold Yjz, it indicates that the water level and flow values ​​in the current water affairs data are relatively dangerous, and the system generates a water affairs warning signal;

[0098] If the water affairs warning coefficient Yj is less than the water affairs warning coefficient threshold Yjz, it means that the water level and flow values ​​in the current water affairs data are within the warning range, but it is still necessary to continuously monitor the changes in the water affairs warning coefficient Yj to generate AI analysis signals;

[0099] It should be noted that the staff receives the warning signal and takes early warning measures to manage the warning; generates AI analysis signals and uses artificial intelligence algorithms to perform predictive analysis on water data;

[0100] Step 4: Based on AI analysis signals, historical water affairs data and meteorological data are obtained to build a smart water conservancy model. The model is used to predict and calculate the future water affairs warning coefficient Yj to obtain the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it indicates that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated;

[0101] Obtain historical water affairs data, including: water affairs data of multiple monitoring periods, water affairs early warning coefficient Yj, and meteorological data corresponding to the monitoring period, where meteorological data includes: rainfall and temperature;

[0102] Based on historical water affairs data and meteorological data, a smart water conservancy model is constructed using the recurrent neural network algorithm (RNN) in the field of artificial intelligence;

[0103] Specifically, the construction method of the smart water conservancy model is as follows:

[0104] SS1, based on the historical water affairs data and meteorological data obtained, the model data set is integrated, and the data set is preprocessed to remove outliers and missing values;

[0105] SS2, using the recurrent neural network (RNN) architecture and ReLU as the activation function, the preprocessed data set is divided into a test set and a training set;

[0106] SS3, use the training set to train the model, and use the validation set to evaluate the model training results;

[0107] Based on the constructed smart water conservancy model, water conservancy data is predicted and analyzed according to future meteorological data;

[0108] Specifically, input the future monitoring period data and future meteorological data, use the smart water conservancy model to calculate the future water conservancy warning coefficient Yj, and obtain the water conservancy prediction coefficient Yc;

[0109] If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it means that the water affairs prediction coefficient Yc exceeds expectations, and a prediction warning signal is generated. After receiving the prediction warning signal, the relevant personnel take necessary measures to eliminate the warning;

[0110] If the water affairs forecast coefficient Yc is less than the water affairs early warning coefficient threshold Yjz, it means that the current water affairs forecast coefficient is still within the expected range and it is still necessary to continuously monitor the changes in the water affairs forecast coefficient Yc;

[0111] The technical solution of this embodiment is: based on the water affairs analysis signal, use the sliding window method to analyze the continuity of the water affairs analysis signal, obtain continuous monitoring moments, calculate the continuous moment ratio Lx, obtain the difference between the continuously monitored water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz, calculate the comprehensive mean ratio Zh, calculate the water affairs early warning coefficient Yj based on the continuous moment ratio Lx and the comprehensive mean ratio Zh, the water affairs early warning coefficient Yj<water affairs early warning coefficient threshold Yjz, generate AI analysis signal, obtain historical water affairs data and meteorological data, build a smart water conservancy model, use the model to predict and calculate the future water affairs early warning coefficient Yj, and obtain the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc≥water affairs early warning coefficient threshold Yjz, it means that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated.

[0112] Example 3

[0113] like Figure 2 As shown, the present invention is a smart water affairs and water conservancy algorithm development system based on artificial intelligence, comprising:

[0114] Data acquisition module: collects water affairs data and stores it in the database;

[0115] Data analysis module: obtain the water level value and flow value of the water affairs data in the database at each monitoring moment during the monitoring period, calculate the water level value and flow value with the corresponding water level warning value and flow warning value respectively, obtain the water level warning ratio Sw and flow warning ratio Ll, further calculate the water affairs analysis coefficient Fx, and generate a water affairs analysis signal if the water affairs analysis coefficient Fx ≥ the water affairs analysis threshold Fxz;

[0116] Early warning judgment module: Based on the water affairs analysis signal, the sliding window method is used to analyze the continuity of the water affairs analysis signal, obtain the continuous monitoring time, calculate the continuous time ratio Lx, obtain the difference between the continuously monitored water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz, calculate the comprehensive mean ratio Zh, and calculate the water affairs early warning coefficient Yj based on the continuous time ratio Lx and the comprehensive mean ratio Zh. The water affairs early warning coefficient Yj < the water affairs early warning coefficient threshold Yjz, and generate an AI analysis signal;

[0117] AI prediction module: Based on AI analysis signals, historical water affairs data and meteorological data are obtained to build a smart water conservancy model. The model is used to predict and calculate the future water affairs warning coefficient Yj to obtain the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it indicates that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated.

[0118] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for developing smart water affairs and water conservancy algorithms based on artificial intelligence, characterized in that: Collect water service data and store it in the database; Obtain the water level value and flow value in the database at each monitoring moment during the monitoring period, calculate the water level value and flow value with the corresponding water level warning value and flow warning value respectively, and obtain the water level warning ratio Sw and flow warning ratio Ll, and further calculate the water affairs analysis coefficient Fx. If the water affairs analysis coefficient Fx ≥ the water affairs analysis threshold Fxz, generate a water affairs analysis signal; Based on the water affairs analysis signal, the sliding window method is used to analyze the continuity of the water affairs analysis signal, obtain continuous monitoring moments, calculate the continuous moment ratio Lx, obtain the difference between the continuously monitored water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz, calculate the comprehensive mean ratio Zh, and calculate the water affairs early warning coefficient Yj based on the continuous moment ratio Lx and the comprehensive mean ratio Zh. The water affairs early warning coefficient Yj < the water affairs early warning coefficient threshold Yjz, and generate an AI analysis signal; Based on AI analysis signals, historical water affairs data and meteorological data are obtained to build a smart water conservancy model. The model is used to predict and calculate the future water affairs warning coefficient Yj to obtain the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it means that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated; The water service analysis coefficient Fx is obtained as follows: Based on the water level warning ratio Sw and flow warning ratio Ll, the water affairs analysis coefficient Fx is calculated through a weighted formula; Compare the water service analysis coefficient Fx with the water service analysis threshold Fxz to generate a water service analysis signal; The water level warning ratio Sw is obtained as follows: Obtain the water level value in the database at each monitoring moment during the monitoring period, make a difference between the water level value and the water level warning value, and take the absolute value to obtain the water level warning difference; The water level warning difference is processed as a ratio with the water level warning value to obtain the water level warning ratio, and the water level warning ratio is marked as Sw; The flow alert ratio L1 is obtained as follows: Obtain the flow value in the database at each monitoring moment during the monitoring period, make a difference between the flow value and the flow warning value, and take the absolute value to obtain the flow warning difference; The flow warning difference is processed as a ratio with the flow warning value to obtain the flow warning ratio, and the flow warning ratio is marked as L1; The water service analysis signal is generated in the following manner: if the water service analysis coefficient Fx≥water service analysis threshold Fxz, a water service analysis signal is generated; The water affairs early warning coefficient Yj is obtained as follows: Based on the continuous time ratio Lx and the comprehensive mean ratio Zh, the water affairs early warning coefficient Yj is calculated; By formula: , the water affairs early warning coefficient Yj is calculated, where ln(b1*Lx+b2*Zh+1.011) is a logarithmic function with base e, and b1 and b2 are preset proportional coefficients; The continuous time ratio Lx is obtained as follows: Get the size of the sliding window, that is, the number of continuous moments, each time the sliding window is paused, and sum up all the continuous moment times to obtain the continuous moment sum; Obtain the number of all monitoring moments within the monitoring period, perform ratio processing on the sum of the continuous moments and the number of all monitoring moments to obtain the continuous moment ratio, and mark the continuous moment ratio as Lx; The method for obtaining the comprehensive mean ratio Zh is: Obtain the difference between the water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz at each moment in the sliding window each time the sliding window is paused, and obtain the analysis coefficient difference; The analysis coefficient differences when the sliding window is paused each time are summed and averaged to obtain the coefficient difference mean; During the monitoring period, all coefficient difference means are summed and averaged to obtain the comprehensive coefficient difference mean; The comprehensive coefficient difference mean is processed by ratio processing with the water affairs analysis coefficient threshold Fxz to obtain the comprehensive mean ratio, which is marked as Zh; The steps of analyzing the continuity of the water service analysis signal with the sliding window are as follows: S1. Determine two consecutive monitoring moments as a sliding window, and place the sliding window at the starting position of the monitoring cycle, that is, the first monitoring moment; S2, obtaining a water service analysis signal within the sliding window. If a water service analysis signal is generated at each monitoring moment within the sliding window, it is considered that the water service analysis signal is continuous within the sliding window; S3, move the sliding window forward by one monitoring time, repeat S2, and judge the continuity of the water service analysis signal at the remaining monitoring time. If the sliding window moves to the next monitoring time and no water service analysis signal is generated, suspend monitoring; S4, obtaining the monitoring time and water service analysis coefficient of the sliding window of suspended monitoring, continuing to move the sliding window forward by one monitoring time, and repeating S2 and S3 until the sliding window covers the entire monitoring cycle; The construction method of the smart water conservancy model is as follows: SS1, based on the historical water affairs data and meteorological data obtained, the model data set is integrated, and the data set is preprocessed to remove outliers and missing values; SS2, using a recurrent neural network architecture and ReLU as the activation function, the preprocessed data set is divided into a test set and a training set; SS3. Use the training set to train the model and use the validation set to evaluate the model training results. The prediction warning signal is generated as follows: Based on the constructed smart water conservancy model, water conservancy data is predicted and analyzed according to future meteorological data; Input future monitoring cycle data and future meteorological data, use the smart water conservancy model to calculate the future water conservancy early warning coefficient Yj, and obtain the water conservancy prediction coefficient Yc; If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it means that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated.

2. A smart water and water conservancy algorithm development system based on artificial intelligence, the system implements the smart water and water conservancy algorithm development method based on artificial intelligence as claimed in claim 1, characterized in that: include: Data acquisition module: collects water affairs data and stores it in the database; Data analysis module: obtain the water level value and flow value in the database at each monitoring moment in the monitoring period, calculate the water level value and flow value with the corresponding water level warning value and flow warning value respectively, obtain the water level warning ratio Sw and flow warning ratio Ll, further calculate the water affairs analysis coefficient Fx, if the water affairs analysis coefficient Fx ≥ the water affairs analysis threshold Fxz, generate a water affairs analysis signal; Early warning judgment module: Based on the water affairs analysis signal, the sliding window method is used to analyze the continuity of the water affairs analysis signal, obtain the continuous monitoring time, calculate the continuous time ratio Lx, obtain the difference between the continuously monitored water affairs analysis coefficient Fx and the water affairs analysis coefficient threshold Fxz, calculate the comprehensive mean ratio Zh, and calculate the water affairs early warning coefficient Yj based on the continuous time ratio Lx and the comprehensive mean ratio Zh. The water affairs early warning coefficient Yj < the water affairs early warning coefficient threshold Yjz, and generate an AI analysis signal; AI prediction module: Based on AI analysis signals, historical water affairs data and meteorological data are obtained to build a smart water conservancy model. The model is used to predict and calculate the future water affairs warning coefficient Yj to obtain the water affairs prediction coefficient Yc. If the water affairs prediction coefficient Yc ≥ the water affairs warning coefficient threshold Yjz, it indicates that the water affairs prediction coefficient Yc exceeds expectations and a prediction warning signal is generated.

Images