Water quality monitoring system based on Internet of Things

Abstract

The invention provides a water quality monitoring system based on the Internet of Things, and the system comprises a sensor network which is used for obtaining related parameter data of a target water area, and transmitting the parameter data to a data processing and analysis center in real time through an Internet of Things communication module; the data processing and analysis center is used for processing and analyzing the parameter data to obtain prediction data and sending the prediction data to the user interaction platform; and the user interaction platform is used for visually displaying the prediction data. According to the invention, real-time monitoring of water quality parameters can be realized, and timeliness and accuracy of monitoring can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a water quality monitoring system based on the Internet of Things. Background Art

[0002] Water quality monitoring is crucial for water resource management, environmental protection and public health. Traditional water quality monitoring methods usually rely on manual sampling and laboratory analysis, which have problems such as long monitoring cycle, poor data real-time performance and high monitoring cost. With the rapid development of Internet of Things technology, it is applied to the field of water quality monitoring, making it possible to achieve remote, real-time and continuous water quality monitoring. Summary of the invention

[0003] In view of this, in order to solve the above problems in the prior art, the present invention proposes a water quality monitoring system based on the Internet of Things.

[0004] The present invention solves the above problems by the following technical means:

[0005] A water quality monitoring system based on the Internet of Things, characterized by comprising:

[0006] Sensor networks, IoT communication modules, data processing and analysis centers, user interaction platforms, and power management systems;

[0007] The sensor network is connected to the Internet of Things communication module, the Internet of Things communication module is connected to the data processing and analysis center, the data processing and analysis center is connected to the user interaction platform, and the power management system is connected to the sensor network and the Internet of Things communication module to perform power supply management on them, wherein:

[0008] The sensor network is used to obtain relevant parameter data of the target water area, and send the parameter data to the data processing and analysis center in real time through the Internet of Things communication module;

[0009] The data processing and analysis center is used to process and analyze the parameter data to obtain prediction data, and send the prediction data to the user interaction platform;

[0010] The user interaction platform is used to visualize the prediction data.

[0011] Furthermore, the sensor network includes several groups of sensors and an underwater robot equipped with a collector, wherein each group of the sensors is deployed at different locations of the target waters according to a preset strategy to obtain relevant parameter data of the target waters, and the underwater robot moves underwater according to a predetermined time period and a planned path to achieve communication connection between the collector and the sensors and collect parameter data of sensors located at different locations.

[0012] Furthermore, the sensor network includes several groups of sensors and underwater robots equipped with collectors, specifically:

[0013] The target water area is divided into a number of grids of the same specification, and a group of sensors are arranged in each grid to obtain a first sensor network structure;

[0014] Optimizing and adjusting the positions of each group of sensors through a preset algorithm to obtain a second sensor network structure, and deploying sensors according to the second sensor network structure;

[0015] According to the self-organizing map, a first path from the underwater robot to each group of sensors for collecting water quality data is obtained;

[0016] The first path is optimized according to a preset strategy to obtain a second path, and according to the second path and a preset time period, the underwater robot is controlled to perform underwater activities to collect relevant water quality parameter data collected by each group of sensors.

[0017] Furthermore, the positions of each group of sensors are optimized and adjusted by a preset algorithm to obtain a second sensor network structure, and sensors are deployed according to the second sensor network structure, wherein the preset algorithm is a genetic algorithm, and specifically includes the following steps:

[0018] Step 1, initializing parameters of the first sensor network structure;

[0019] Step 2, calculating the fitness function of the individuals in the genetic algorithm population, wherein the coverage rate of the sensor is used as the fitness function, and calculating whether the individual meets the coverage rate. The fitness function of the individual that meets the requirements remains unchanged; the fitness function of the individual that does not meet the requirements is modified accordingly;

[0020] Step 3: Sort the individuals according to the size of the fitness function, select the corresponding individuals for crossover, put the new individuals into the next generation population, and finally perform mutation operation on the population;

[0021] Step 4, check whether the current iteration number reaches the preset maximum iteration number, if so, terminate and use the current population as the second sensor network structure, otherwise jump to step 2.

[0022] Further, the first path is optimized according to the preset strategy to obtain the second path, and the underwater robot is controlled to perform underwater activities according to the second path and the preset time period to collect relevant water quality parameter data collected by each group of sensors, wherein the first path is optimized according to the preset strategy to obtain the second path, specifically including the following contents:

[0023] The first path is converted into a 1×1 two-dimensional coordinate system, and a graphic analysis is performed on the first path to determine the initial cluster center of the K-means algorithm;

[0024] Divide the initial cluster centers according to the Euclidean distance to obtain the final cluster centers;

[0025] The coordinate position of the collection point is calculated based on the final cluster center and the preset formula;

[0026] The coordinate positions of the collected points are serially connected in sequence to obtain the second path.

[0027] Furthermore, each group of sensors includes multiple water quality data sensors with different properties, specifically, a temperature sensor for detecting water temperature, a pH meter for measuring pH value, a dissolved oxygen meter for detecting dissolved oxygen content, a turbidity meter for detecting turbidity, a conductivity meter for detecting conductivity, an ammonia nitrogen tester for detecting ammonia nitrogen content, an ammonia nitrogen tester for detecting total ammonia content, a heavy metal content detector for detecting heavy metal content and / or a microbial indicator detector for detecting microorganisms.

[0028] Furthermore, the data processing and analysis center includes:

[0029] A processing module, used for filling missing data in the parameter data by an average value filling method to obtain repair data, and performing feature selection based on the repair data to obtain feature data;

[0030] An optimization module is used to optimize the feature data to obtain optimized data;

[0031] A prediction module is used to predict water quality status based on the optimized data to obtain prediction data;

[0032] The communication module is used to send the prediction data to the user interaction platform.

[0033] Furthermore, the optimization module is used to optimize the feature data by using a sparrow search algorithm to obtain optimized data, which specifically includes the following steps:

[0034] Step 1, initializing and setting the population size of the sparrow search algorithm to a predetermined value, initializing the position of the individual, the population boundary and the maximum number of iterations, and calculating the fitness value;

[0035] Step 2, update the position of the individual according to the preset fitness function;

[0036] Step 3, determine whether the termination condition is met, the termination condition is that the maximum number of iterations is reached or the fitness value reaches the set threshold. If it is met, the optimal solution is obtained and the optimal solution is output as the optimization data, otherwise return to step 1.

[0037] Furthermore, the prediction module includes:

[0038] Building blocks, specifically for:

[0039] Obtain the data released by the national surface water quality automatic monitoring station, and select at least 500 water quality data from the released data as training sample data, and count the water quality categories as c i The number of instances of COUNT(c i ) and under this category attribute A j The value is a j Number of instances of COUNT(A j =a j ^c i ), forming an instance statistics table;

[0040] According to the instance statistics table, calculate the prior probability P(c i ) and in category c i Attribute A j The value is a j The prior probability P(a j |c i ), save the result in the prior probability table;

[0041] According to the results of the prior probability table, a naive Bayes classification is performed on each attribute, and the weight w of each attribute is calculated based on the probability of successful classification of each attribute. 1，j , forming a weight table;

[0042] The prediction module is used to calculate the weight w of each attribute based on the optimized data 2，j ;

[0043] According to the attribute weight w 2，j The approximation comparison is performed in combination with the weight table, and the predicted data is obtained through the comparison results.

[0044] Compared with the prior art, the beneficial effects of the present invention include at least:

[0045] 1. The present invention realizes real-time monitoring of water quality parameters through the Internet of Things technology, which can effectively improve the timeliness and accuracy of monitoring. In addition, since a data processing and analysis center is set up to process and analyze water quality parameter data to obtain predicted data, water quality problems can be discovered in advance based on the predicted data.

[0046] 2. Since the present invention provides a user interaction platform to visualize the predicted data, it is convenient for users to remotely and intuitively understand or monitor water quality data, and can effectively reduce labor costs.

[0047] 3. The present invention can effectively reduce power consumption and extend the service life of the system by setting up a power management system to manage the power supply of the sensor network and the Internet of Things communication module. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 It is a schematic diagram of the structure of a water quality monitoring system based on the Internet of Things provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the technical solutions of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] like Figure 1 As shown, an embodiment of the present invention provides a water quality monitoring system based on the Internet of Things, including:

[0052] Sensor network 1, IoT communication module 2, data processing and analysis center 3, user interaction platform 4 and power management system 5;

[0053] The sensor network 1 is connected to the Internet of Things communication module 2, the Internet of Things communication module 2 is connected to the data processing and analysis center 3, the data processing and analysis center 3 is connected to the user interaction platform 4, and the power management system 5 is connected to the sensor network 1 and the Internet of Things communication module 2 to manage their power supply, wherein:

[0054] The sensor network 1 is used to obtain relevant parameter data of the target water area, and send the parameter data to the data processing and analysis center 3 in real time through the Internet of Things communication module 2;

[0055] The data processing and analysis center 3 is used to process and analyze the parameter data to obtain prediction data, and send the prediction data to the user interaction platform 4;

[0056] The user interaction platform 4 is used to visualize the prediction data.

[0057] Obviously, the present invention realizes real-time monitoring of water quality parameters through the Internet of Things technology, which can effectively improve the timeliness and accuracy of monitoring. In addition, since a data processing and analysis center is set up to process and analyze water quality parameter data to obtain predicted data, water quality problems can be discovered in advance based on the predicted data.

[0058] At the same time, since a user interaction platform is set up to visualize the predicted data, it is convenient for users to remotely and intuitively understand or monitor water quality data, and it can effectively reduce labor costs.

[0059] In addition, by setting up a power management system to manage the power supply of the sensor network and IoT communication modules, power consumption can be effectively reduced and the service life of the system can be extended.

[0060] In an embodiment of the present invention, the sensor network includes several groups of sensors and an underwater robot equipped with a collector, wherein each group of sensors is deployed at different locations of the target water area according to a preset strategy to obtain relevant parameter data of the target water area, and the underwater robot moves underwater according to a predetermined time period and a planned path to achieve communication connection between the collector and the sensors and collect parameter data of sensors located at different locations.

[0061] Because the target water area usually has a large span and a wide range, it is necessary to set up sensors at different locations to achieve comprehensive and accurate monitoring of the water quality of the target water area. However, since the sensors at different locations are too far apart, it is difficult to collect unified water quality parameter data. Therefore, setting up an underwater robot equipped with a collector to collect water quality parameter data can perfectly solve this problem.

[0062] At the same time, since the underwater robot is set to move underwater according to a predetermined time period and planned path to realize the communication connection between the collector and the sensor and collect parameter data of sensors located at different positions, it can save energy and cost and avoid missed collection or repeated collection.

[0063] In some more specific embodiments, the above scheme may include the following contents:

[0064] The target water area is divided into a number of grids of the same specification, and a group of sensors are arranged in each grid to obtain a first sensor network structure;

[0065] Optimizing and adjusting the positions of each group of sensors through a preset algorithm to obtain a second sensor network structure, and deploying sensors according to the second sensor network structure;

[0066] According to the self-organizing map, a first path from the underwater robot to each group of sensors for collecting water quality data is obtained;

[0067] The first path is optimized according to a preset strategy to obtain a second path, and according to the second path and a preset time period, the underwater robot is controlled to perform underwater activities to collect relevant water quality parameter data collected by each group of sensors.

[0068] It can be seen that since the target water area is divided into several grids of the same specifications and a group of sensors are set in each grid, comprehensive and complete monitoring of the target water area can be achieved, and since the positions of each group of sensors are optimized and adjusted by a preset algorithm to obtain the second sensor network structure, and sensors are deployed according to the second sensor network structure, the performance of the sensor network structure is effectively improved, making it more reasonable.

[0069] At the same time, since the first path from the underwater robot to each group of sensors for collecting water quality data is obtained according to the self-organizing map; the first path is optimized according to the preset strategy to obtain the second path, and according to the second path and the preset time period, the underwater robot is controlled to perform underwater activities to collect relevant water quality parameter data collected by each group of sensors, therefore, the underwater robot can perform effective data collection, avoid missed collection or repeated collection of water quality parameter data, and improve the degree of intelligence.

[0070] For example, the positions of each group of sensors are optimized and adjusted by a preset algorithm to obtain a second sensor network structure, and sensors are deployed according to the second sensor network structure, wherein the preset algorithm is a genetic algorithm, and specifically may include the following steps:

[0071] Step 1, initializing the parameters of the first sensor network structure, including the number of grids, the distance between grid points, the grid type, the sensor coverage range and other parameters, and initializing the relevant parameters involved in the genetic algorithm, including the selection function, mutation probability and crossover method of the genetic algorithm;

[0072] Step 2, calculating the fitness function of the individuals in the genetic algorithm population, wherein the coverage rate of the sensor is used as the fitness function, and calculating whether the individual meets the coverage rate. The fitness function of the individual that meets the requirements remains unchanged; the fitness function of the individual that does not meet the requirements is modified accordingly;

[0073] Step 3: Sort the individuals according to the size of the fitness function, select the corresponding individuals for crossover, put the new individuals into the next generation population, and finally perform mutation operation on the population;

[0074] Step 4, check whether the current iteration number reaches the preset maximum iteration number, if so, terminate and use the current population as the second sensor network structure, otherwise jump to step 2.

[0075] Obviously, since the genetic algorithm has the characteristic of rapid convergence, it can reduce the amount of calculation and the calculation time. Therefore, the optimal second sensor network structure can be quickly obtained through the above algorithm.

[0076] For example, the first path is optimized according to a preset strategy to obtain a second path, and according to the second path and a preset time period, the underwater robot is controlled to perform underwater activities to collect relevant water quality parameter data collected by each group of sensors. The first path is optimized according to the preset strategy to obtain the second path, which may specifically include the following contents:

[0077] The first path is converted into a 1×1 two-dimensional coordinate system, and a graphic analysis is performed on the first path to determine the initial cluster center of the K-means algorithm;

[0078] Divide the initial cluster centers according to the Euclidean distance to obtain the final cluster centers;

[0079] The coordinate position of the collection point is calculated based on the final cluster center and the preset formula;

[0080] The coordinate positions of the collected points are serially connected in sequence to obtain the second path.

[0081] Obviously, since the first path is optimized by the improved K-means algorithm, a better clustering effect can be achieved, making the second path more reasonable.

[0082] For example, the coordinate position of the collection point is calculated according to the final cluster center and the preset formula, wherein the preset formula can be expressed by the following formula:

[0083]

[0084] In the above formula, U k (u kr ,u ky ) represents the coordinates of the center of the k-th cluster, C(c x , c y ) represents the coordinates of the sensor node belonging to the center of the kth cluster, M(m x , m y ) represents the coordinates of the collection point.

[0085] In addition, each group of sensors includes multiple water quality data sensors with different properties, for example, a temperature sensor for detecting water temperature, a pH meter for measuring pH value, a dissolved oxygen meter for detecting dissolved oxygen content, a turbidity meter for detecting turbidity, a conductivity meter for detecting conductivity, an ammonia nitrogen tester for detecting ammonia nitrogen content, an ammonia nitrogen tester for detecting total ammonia content, a heavy metal content detector for detecting heavy metal content, and a microbial index detector for detecting microorganisms, etc., which can be specifically configured as needed and are not limited here;

[0086] And each sensor adopts low power consumption design to ensure long-term stable operation.

[0087] In the embodiment of the present invention, the data processing and analysis center includes:

[0088] A processing module, used for filling missing data in the parameter data by an average value filling method to obtain repair data, and performing feature selection based on the repair data to obtain feature data;

[0089] An optimization module is used to optimize the feature data to obtain optimized data;

[0090] A prediction module is used to predict water quality status based on the optimized data to obtain prediction data;

[0091] The communication module is used to send the prediction data to the user interaction platform.

[0092] Obviously, the water quality parameter data collected by each sensor will inevitably have missing data due to faults, surrounding environment, etc. In order to improve the prediction accuracy, the data needs to be filled and preprocessed. The embodiment of the present invention fills the missing data in the parameter data to obtain repair data, thereby ensuring the integrity of the data and thus improving the reliability of the prediction results. Moreover, since feature data is obtained by feature selection based on the repair data, the complexity of the problem can be reduced and the computing cost can be saved.

[0093] For example, the optimization module is used to optimize the feature data by using the sparrow search algorithm to obtain optimized data, which may specifically include the following steps:

[0094] Step 1, initializing and setting the population size of the sparrow search algorithm to a predetermined value, initializing the position of the individual, the population boundary and the maximum number of iterations, and calculating the fitness value;

[0095] Step 2, update the position of the individual according to the preset fitness function;

[0096] Step 3, determine whether the termination condition is met, the termination condition is that the maximum number of iterations is reached or the fitness value reaches the set threshold. If it is met, the optimal solution is obtained and the optimal solution is output as the optimization data, otherwise return to step 1.

[0097] An example, prediction module, includes:

[0098] Building modules are used to construct classifiers, which can be:

[0099] Obtain the data released by the national surface water quality automatic monitoring station, and select at least 500 water quality data from the released data as training sample data, and count the water quality categories as c i The number of instances of COUNT(c i ) and under this category attribute A j The value is a jNumber of instances of COUNT(A j =A j ^c i ), forming an instance statistics table;

[0100] According to the instance statistics table, calculate the prior probability P(c i ) and in category c i Attribute A j The value is a j The prior probability P(a j |c i ), save the result in the prior probability table;

[0101] According to the results of the prior probability table, a naive Bayes classification is performed on each attribute, and the weight w of each attribute is calculated based on the probability of successful classification of each attribute. 1，j , forming a weight table;

[0102] For example, the weight table is as follows:

[0103]

[0104] The prediction module is used to calculate the weight w of each attribute based on the optimized data 2，j ;

[0105] According to the attribute weight w 2，j The approximation comparison is performed in combination with the weight table, and the predicted data is obtained through the comparison results.

[0106] Obviously, the embodiment of the present invention obtains a weight table through training sample data, performs calculations based on optimized data, and performs data prediction in a table lookup manner, which can effectively improve accuracy and precision.

[0107] Although the description of the present invention has been quite detailed and specifically describes several described embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, but should be regarded as providing a broad possible interpretation of these claims by reference to the appended claims, taking into account the prior art, so as to effectively cover the intended scope of the present invention. In addition, the above description of the present invention is based on the embodiments foreseeable by the inventor, and its purpose is to provide a useful description, and those non-substantial changes to the present invention that have not yet been foreseen may still represent equivalent changes to the present invention.

Claims

1. A water quality monitoring system based on the Internet of Things, characterized in that: include: Sensor networks, IoT communication modules, data processing and analysis centers, user interaction platforms, and power management systems; The sensor network is connected to the Internet of Things communication module, the Internet of Things communication module is connected to the data processing and analysis center, the data processing and analysis center is connected to the user interaction platform, and the power management system is connected to the sensor network and the Internet of Things communication module to perform power supply management on them, wherein: The sensor network is used to obtain relevant parameter data of the target water area, and send the parameter data to the data processing and analysis center in real time through the Internet of Things communication module; The data processing and analysis center is used to process and analyze the parameter data to obtain prediction data, and send the prediction data to the user interaction platform; The user interaction platform is used to visualize the prediction data.

2. The water quality monitoring system based on the Internet of Things according to claim 1 is characterized in that: The sensor network includes several groups of sensors and an underwater robot equipped with a collector, wherein each group of the sensors is deployed at different locations of the target water area according to a preset strategy to obtain relevant parameter data of the target water area, and the underwater robot moves underwater according to a predetermined time period and a planned path to achieve communication connection between the collector and the sensors and collect parameter data of sensors located at different locations.

3. The water quality monitoring system based on the Internet of Things according to claim 1 is characterized in that: The sensor network includes several groups of sensors and underwater robots equipped with collectors, specifically: The target water area is divided into a number of grids of the same specification, and a group of sensors are arranged in each grid to obtain a first sensor network structure; Optimizing and adjusting the positions of each group of sensors through a preset algorithm to obtain a second sensor network structure, and deploying sensors according to the second sensor network structure; According to the self-organizing map, a first path from the underwater robot to each group of sensors for collecting water quality data is obtained; The first path is optimized according to a preset strategy to obtain a second path, and according to the second path and a preset time period, the underwater robot is controlled to perform underwater activities to collect relevant water quality parameter data collected by each group of sensors.

4. The water quality monitoring system based on the Internet of Things according to claim 3 is characterized in that: The positions of the sensors of each group are optimized and adjusted by a preset algorithm to obtain a second sensor network structure, and the sensors are deployed according to the second sensor network structure, wherein the preset algorithm is a genetic algorithm, and specifically includes the following steps: Step 1, initializing the parameters of the first sensor network structure; Step 2, calculating the fitness function of the individuals in the genetic algorithm population, wherein the coverage rate of the sensor is used as the fitness function, and calculating whether the individual meets the coverage rate. The fitness function of the individual that meets the requirements remains unchanged; the fitness function of the individual that does not meet the requirements is modified accordingly; Step 3: Sort the individuals according to the size of the fitness function, select the corresponding individuals for crossover, put the new individuals into the next generation population, and finally perform mutation operation on the population; Step 4, check whether the current iteration number reaches the preset maximum iteration number, if so, terminate and use the current population as the second sensor network structure, otherwise jump to step 2.

5. The water quality monitoring system based on the Internet of Things according to claim 3 is characterized in that: The first path is optimized according to the preset strategy to obtain the second path, and the underwater robot is controlled to perform underwater activities according to the second path and the preset time period to collect relevant water quality parameter data collected by each group of sensors, wherein the first path is optimized according to the preset strategy to obtain the second path, specifically including the following contents: The first path is converted into a 1×1 two-dimensional coordinate system, and a graphic analysis is performed on the first path to determine the initial cluster center of the K-means algorithm; Divide the initial cluster centers according to the Euclidean distance to obtain the final cluster centers; The coordinate position of the collection point is calculated based on the final cluster center and the preset formula; The coordinate positions of the collected points are serially connected in sequence to obtain the second path.

6. The water quality monitoring system based on the Internet of Things according to claim 2 is characterized in that: Each group of sensors includes multiple water quality data sensors with different properties, specifically, a temperature sensor for detecting water temperature, a pH meter for measuring pH value, a dissolved oxygen meter for detecting dissolved oxygen content, a turbidity meter for detecting turbidity, a conductivity meter for detecting conductivity, an ammonia nitrogen tester for detecting ammonia nitrogen content, an ammonia nitrogen tester for detecting total ammonia content, a heavy metal content detector for detecting heavy metal content and / or a microbial indicator detector for detecting microorganisms.

7. The water quality monitoring system based on the Internet of Things according to claim 1 is characterized in that: The data processing and analysis center includes: A processing module, used for filling missing data in the parameter data by an average value filling method to obtain repair data, and performing feature selection based on the repair data to obtain feature data; An optimization module is used to optimize the feature data to obtain optimized data; A prediction module is used to predict water quality status based on the optimized data to obtain prediction data; The communication module is used to send the prediction data to the user interaction platform.

8. The water quality monitoring system based on the Internet of Things according to claim 7 is characterized in that: The optimization module is used to optimize the feature data by using the sparrow search algorithm to obtain optimized data, which specifically includes the following steps: Step 1, initializing and setting the population size of the sparrow search algorithm to a predetermined value, initializing the position of the individual, the population boundary and the maximum number of iterations, and calculating the fitness value; Step 2, update the position of the individual according to the preset fitness function; Step 3, determine whether the termination condition is met, the termination condition is that the maximum number of iterations is reached or the fitness value reaches the set threshold. If it is met, the optimal solution is obtained and the optimal solution is output as the optimization data, otherwise return to step 1.

9. The water quality monitoring system based on the Internet of Things according to claim 8, characterized in that: The prediction module comprises: Building blocks, specifically for: Obtain the data released by the national surface water quality automatic monitoring station, and select at least 500 water quality data from the released data as training sample data, and count the water quality categories as c i The number of instances of COUNT(c i ) and under this category attribute A j The value is a j Number of instances of COUNT(A j =a j ^c i ), forming an instance statistics table; According to the instance statistics table, calculate the prior probability P(c i ) and in category c i Attribute A j The value is a j The prior probability P(a j |c i ), save the result in the prior probability table; According to the results of the prior probability table, a naive Bayes classification is performed on each attribute, and the weight w of each attribute is calculated based on the probability of successful classification of each attribute. 1，j , forming a weight table; The prediction module is used to calculate the weight w of each attribute based on the optimized data 2，j ; According to the attribute weight w 2，j The approximation comparison is performed in combination with the weight table, and the predicted data is obtained through the comparison results.

Images