CN111105079A - Internet of things water body dissolved oxygen prediction method and water quality monitoring system - Google Patents

Internet of things water body dissolved oxygen prediction method and water quality monitoring system Download PDF

Info

Publication number
CN111105079A
CN111105079A CN201911202271.4A CN201911202271A CN111105079A CN 111105079 A CN111105079 A CN 111105079A CN 201911202271 A CN201911202271 A CN 201911202271A CN 111105079 A CN111105079 A CN 111105079A
Authority
CN
China
Prior art keywords
module
dissolved oxygen
monitoring system
quality monitoring
water quality
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201911202271.4A
Other languages
Chinese (zh)
Inventor
匡亮
施珮
华驰
邓小龙
顾晓燕
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Vocational College of Information Technology
Original Assignee
Jiangsu Vocational College of Information Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiangsu Vocational College of Information Technology filed Critical Jiangsu Vocational College of Information Technology
Priority to CN201911202271.4A priority Critical patent/CN111105079A/en
Publication of CN111105079A publication Critical patent/CN111105079A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Forestry; Mining

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Biomedical Technology (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Biophysics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Strategic Management (AREA)
  • Artificial Intelligence (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Data Mining & Analysis (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Agronomy & Crop Science (AREA)
  • Animal Husbandry (AREA)
  • Marine Sciences & Fisheries (AREA)
  • Mining & Mineral Resources (AREA)
  • Primary Health Care (AREA)
  • Investigating Or Analyzing Materials By The Use Of Electric Means (AREA)

Abstract

The invention discloses a method for predicting water body dissolved oxygen of an internet of things and a water quality monitoring system, wherein data of air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH, conductivity and dissolved oxygen are obtained through the water quality monitoring system, the data of the air temperature, the humidity, the rainfall, the wind speed, the air pressure, the water temperature, the pH and the conductivity are set as input quantity, the data of the dissolved oxygen are output quantity, and a small habitat thinking evolution algorithm (NMEA) is utilized to optimize a BP neural network to construct an NMEA-BP model to predict the dissolved oxygen; aiming at the problems of easy falling into local minimum value, low prediction precision and the like in the prior art, the method for predicting the dissolved oxygen in the water body of the Internet of things and the water quality monitoring system provided by the invention build the NMEA-BP model to predict the dissolved oxygen, can avoid the problem of falling into the local minimum value, and have high prediction precision.

Description

Internet of things water body dissolved oxygen prediction method and water quality monitoring system
Technical Field
The invention relates to a method for predicting water body dissolved oxygen of the Internet of things and a water quality monitoring system.
Background
Molecular oxygen dissolved in air in water is called dissolved oxygen, and the dissolved oxygen is a necessary condition for fish to live and is a key index for fishermen to stably harvest. The anoxic water body can cause the vitality of the fishes and shrimps to be reduced, the metabolism is slowed down, and the food intake is slowed down. Organic matters, ammonia nitrogen and the like in water are subjected to anaerobic decomposition to generate some toxic substances such as nitrite and the like, and bacteria are easy to breed. When the dissolved oxygen is lower than the minimum limit, the fish can float and even die in a large area by suffocation. Therefore, the real-time detection of the dissolved oxygen content in the water body and the timely prediction of the dissolved oxygen content and the trend are important factors for reducing the risk of aquaculture.
The requirements of different breeding species, different ages and different seasons on the dissolved oxygen of the pond water are different. The content of dissolved oxygen in water is closely related to water environment factors such as meteorological factors and water temperature in the air. At present, researchers at home and abroad propose various methods for predicting dissolved oxygen, a BP neural network or a time series model is used for predicting the content of the dissolved oxygen, a Support Vector Machine (SVM) is used for predicting the content of the dissolved oxygen, a genetic algorithm is used for optimizing the BP neural network for predicting the content of the dissolved oxygen, and the like.
Disclosure of Invention
Aiming at the problems of easy falling into local minimum value and low prediction precision and the like in the prior art, the invention provides a method for predicting water body dissolved oxygen of the Internet of things and a water quality monitoring system.
In order to solve the technical problems, the invention adopts the following technical scheme:
a prediction method for dissolved oxygen in water of the Internet of things comprises the following steps:
the method comprises the following steps: determining a training set and a test set;
acquiring data of air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH, conductivity and dissolved oxygen through a water quality monitoring system, dividing the data into a training set and a testing set, and setting the data of the air temperature, the humidity, the rainfall, the wind speed, the air pressure, the water temperature, the pH and the conductivity as input quantity pnThe data of dissolved oxygen is output Y;
step two: determining a BP neural network topological structure;
input quantity p according to training setnAnd the mapping corresponding relation between the output quantity Y, a three-layer topological structure is established to construct a BP neural network, the number of nodes of an input layer is n, the number of nodes of an output layer is m,
number of hidden layer nodes n1：
Figure BDA0002296154840000021
In the formula, a is an integer between 0 and 10;
the BP neural network model is as follows:
Y＝purelin(W2,1×tansig(W1,1pn+b1)+b2) (2)
wherein Y is the output, W1,1And W2,1Network connection weights from input layer to hidden layer and from hidden layer to output layer, respectively, b1And b2Networks from input layer to hidden layer, hidden layer to output layer, respectivelyConnection threshold, pnFor input quantity, tansig is a transfer function from an input layer to a hidden layer, and purelin is a transfer function from the hidden layer to an output layer;
the method comprises the following steps that a sigmoid function is selected as an activation function between an input layer and an output layer, and the function form is as follows:
Figure BDA0002296154840000022
step three: obtaining the optimal weight and threshold value required by the BP neural network through a niche thought evolution algorithm NMEA;
(a) group initialization
Randomly generating N individuals to form an initial population;
the score of each individual is calculated by the equations (4) and (5),
Figure BDA0002296154840000023
Figure BDA0002296154840000024
in the formula, tiAnd
Figure BDA0002296154840000025
respectively taking the actual value of the ith sample and the output value of the BP neural network model;
(b) generating a new sub-population;
sorting the individuals according to the scores of the individuals from big to small, selecting the first q individuals as winners, and forming sub-groups respectively by taking the winners as centers;
(c) generating a niche;
calculating the distance between the winners of the q sub-groups according to the formula (6), when the distance between the winners of the two sub-groups is smaller than the radius of the niche, discarding the sub-group with low score, keeping the score of the sub-group high, and DijThe distance between two sub-group winners is defined, and s is the individual coding length;
Figure BDA0002296154840000031
(d) determining optima for each niche independent search space
Reinitializing the abandoned sub-population, reselecting a winner in the niche where the abandoned sub-population is located, and turning to the step (c) to judge the niche winner score again until each niche has the winner, namely determining the optimal person of each niche independent search space;
(e) convergent operations of subgroups
Converging the sub-populations, and differentiating the populations with the lowest scores after convergence;
(f) performing diversity operation on the sub-groups to obtain optimal weight values and threshold values;
stopping iteration when the score of the optimal winner does not change any more to obtain an optimal individual, namely obtaining a group of optimal weight values and threshold values;
step four: constructing an NMEA-BP model;
substituting the optimal weight and the threshold obtained in the third step into a formula (2) to construct an NMEA-BP model which is a dissolved oxygen prediction model, wherein the optimal weight and the threshold are used as the initial weight and the threshold of the BP neural network model;
step five: and verifying the NMEA-BP model constructed in the step four according to the test set.
The technical scheme of the invention is further improved, and the water quality monitoring system comprises a collection node, a sink node and a server; the collection node comprises a first microcontroller module, a temperature collection module, a humidity collection module, an air pressure collection module, a rainfall module, an air speed module, a water temperature collection module, a conductivity module, a pH collection module, a dissolved oxygen module, a first LoRa communication module and a first power supply module; the first microcontroller module is respectively connected with the temperature acquisition module, the humidity acquisition module, the air pressure acquisition module, the rainfall module, the air speed module, the water temperature acquisition module, the conductivity module, the pH acquisition module, the dissolved oxygen module and the first LoRa communication module; the first power supply module is connected with other modules in the acquisition node; the sink node comprises a second microcontroller module, a second LoRa communication module, a 4G communication module and a second power supply module; the second microcontroller module is respectively connected with the second LoRa communication module and the 4G communication module; the second power supply module is connected with other modules in the sink node; the collection node sends each module data of collection to the sink node through the loRa network that first loRa communication module and second loRa communication module constitute, the sink node is through 4G network with data transmission to server. The water quality monitoring system provided by the invention works in the 4G network coverage range, greatly facilitates the arrangement of the server, has the advantages of convenient wiring, good flexibility, no limitation of communication distance and the like, and can be widely applied to water quality monitoring of oceans and lakes and the like; wherein the collection node sends data such as temperature, pH value, conductivity and dissolved oxygen to the sink node through the loRa network, and the sink node passes through 4G network transmission to the server, and entire system simple structure, the commonality is good.
According to the further improvement of the technical scheme, the temperature acquisition module and the water temperature acquisition module both adopt DS18B20 sensors which are subjected to waterproof treatment.
The technical scheme of the invention is further improved, and the conductivity module adopts a DJS-1 type platinum black point conductive electrode produced by Shanghai precision scientific instruments Co.
The technical scheme of the invention is further improved, and the pH acquisition module adopts an E-201-C type pH electrode produced by Shanghai precision scientific instruments, Inc.
According to the technical scheme of the invention, the dissolved oxygen acquisition module is a polarographic sensor.
In a further improvement of the technical scheme of the invention, the first microcontroller module and the second microcontroller module both adopt STM32F103RBT6 chips produced by Italian semiconductor corporation.
The technical scheme of the invention is further improved, and is characterized in that: the 4G communication module uses a USIM card and adopts a China mobile network.
By further improving the technical scheme of the invention, a user can access the server to obtain the water quality data through a network at a computer end or a mobile phone end. The water quality data can be acquired at a computer end or a mobile phone end, and the use is convenient.
Compared with the prior art, the invention has the beneficial effects that:
according to the method and the system for predicting the dissolved oxygen in the water body of the Internet of things, provided by the invention, through analyzing the relation between the dissolved oxygen and other monitoring indexes and meteorological indexes of the water body, the data of air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH, conductivity and dissolved oxygen measured by the water quality monitoring system are utilized, the data of air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH and conductivity are set as input quantities, the data of dissolved oxygen is set as output quantities, a small habitat thinking evolution algorithm (NMEA) is used for optimizing a BP neural network to construct an NMEA-BP model to predict the dissolved oxygen, the problem of falling into a local minimum value can be avoided, and the prediction precision is high.
Drawings
FIG. 1 is a flow chart of the prediction method of the present invention.
Fig. 2 is a hardware structure diagram of a water quality monitoring system acquisition node of the invention.
Fig. 3 is a hardware structure diagram of a sink node of the water quality monitoring system of the present invention.
FIG. 4 is a diagram of the overall system architecture of the water quality monitoring system of the present invention.
FIG. 5 is a graph comparing the predicted output value of dissolved oxygen with the true value in the prediction method of the present invention.
Detailed Description
The invention will be described in further detail with reference to the following figures and specific embodiments.
As shown in fig. 1, the prediction method includes the following steps:
the method comprises the following steps: determining a training set and a test set;
acquiring data of air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH, conductivity and dissolved oxygen through a water quality monitoring system, dividing the data into a training set and a testing set, and setting the data of the air temperature, the humidity, the rainfall, the wind speed, the air pressure, the water temperature, the pH and the conductivity as input quantity pnThe data of dissolved oxygen is output Y;
step two: determining a BP neural network topological structure;
input quantity p according to training setnAnd the mapping corresponding relation between the output quantity Y, a three-layer topological structure is established to construct a BP neural network, the number of nodes of an input layer is n, the number of nodes of an output layer is m,
number of hidden layer nodes n1：
Figure BDA0002296154840000051
In the formula, a is an integer between 0 and 10;
the BP neural network model is as follows:
Y＝purelin(W2,1×tansig(W1,1pn+b1)+b2) (2)
wherein Y is the output, W1,1And W2,1Network connection weights from input layer to hidden layer and from hidden layer to output layer, respectively, b1And b2Network connection threshold, p, from input layer to hidden layer, hidden layer to output layer, respectivelynFor input quantity, tansig is a transfer function from an input layer to a hidden layer, and purelin is a transfer function from the hidden layer to an output layer;
the method comprises the following steps that a sigmoid function is selected as an activation function between an input layer and an output layer, and the function form is as follows:
Figure BDA0002296154840000052
step three: obtaining the optimal weight and threshold value required by the BP neural network through a niche thought evolution algorithm NMEA;
(a) group initialization
Randomly generating N individuals to form an initial population;
the score of each individual is calculated by the equations (4) and (5),
Figure BDA0002296154840000061
Figure BDA0002296154840000062
in the formula, tiAnd
Figure BDA0002296154840000063
respectively taking the actual value of the ith sample and the output value of the BP neural network model;
(b) generating a new sub-population;
sorting the individuals according to the scores of the individuals from big to small, selecting the first q individuals as winners, and forming sub-groups respectively by taking the winners as centers;
(c) generating a niche;
calculating the distance between the winners of the q sub-groups according to the formula (6), when the distance between the winners of the two sub-groups is smaller than the radius of the niche, discarding the sub-group with low score, keeping the score of the sub-group high, and DijThe distance between two sub-group winners is defined, and s is the individual coding length;
Figure BDA0002296154840000064
(d) determining optima for each niche independent search space
Reinitializing the abandoned sub-population, reselecting a winner in the niche where the abandoned sub-population is located, and turning to the step (c) to judge the niche winner score again until each niche has the winner, namely determining the optimal person of each niche independent search space;
(e) convergent operations of subgroups
Converging the sub-populations, and differentiating the populations with the lowest scores after convergence;
(f) performing diversity operation on the sub-groups to obtain optimal weight values and threshold values;
stopping iteration when the score of the optimal winner does not change any more to obtain an optimal individual, namely obtaining a group of optimal weight values and threshold values;
step four: constructing an NMEA-BP model;
substituting the optimal weight and the threshold obtained in the third step into a formula (2) to construct an NMEA-BP model which is a dissolved oxygen prediction model, wherein the optimal weight and the threshold are used as the initial weight and the threshold of the BP neural network model;
step five: and verifying the NMEA-BP model constructed in the step four according to the test set.
In the embodiment, data of air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH and conductivity are input quantity, output quantity is dissolved oxygen, the number of nodes of an input layer is determined to be 8, the number of nodes of an output layer is 1, the number of nodes of a hidden layer is determined to be an integer between 3 and 13 after calculation according to the formula (1), after the 11 numbers are tested one by one, the number of nodes of the hidden layer is finally determined to be 9, and the maximum training time is 2000;
in the third step, the optimal weight and the threshold are obtained through a niche thought evolution algorithm NMEA, the initial population N is set to be 200, the dominant sub-population and the temporary sub-population are respectively set to be 5, the niche radius is 20, and the iteration number is set to be 10.
As shown in fig. 2, the collection node in the water quality monitoring system includes a first microcontroller module, a temperature collection module, a humidity collection module, an air pressure collection module, a rainfall module, an air speed module, a water temperature collection module, a conductivity module, a pH collection module, a dissolved oxygen module, a first LoRa communication module and a first power supply module;
the first microcontroller module is respectively connected with the temperature acquisition module, the humidity acquisition module, the air pressure acquisition module, the rainfall module, the air speed module, the water temperature acquisition module, the conductivity module, the pH acquisition module, the dissolved oxygen module and the first LoRa communication module;
the first power supply module is connected with other modules in the acquisition node;
wherein first microcontroller module is used for reading the data that temperature acquisition module, humidity acquisition module, atmospheric pressure acquisition module, rainfall module, wind speed module, temperature acquisition module, conductivity module, pH acquisition module and dissolved oxygen module gathered to send data for first loRa communication module.
As shown in fig. 3, the sink node in the water quality monitoring system includes a second microcontroller module, a second LoRa communication module, a 4G communication module, and a second power module;
the second microcontroller module is respectively connected with the second LoRa communication module and the 4G communication module;
the second power supply module is connected with other modules in the sink node;
the second LoRa communication module sends data to the second microcontroller module, and the second microcontroller module sends the data to the through 4G communication module.
As shown in fig. 4, the collection node in the water quality monitoring system sends the water quality data air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH value, conductivity and dissolved oxygen to the sink node through the LoRa network formed by the first LoRa communication module and the second LoRa communication module, the sink node transmits the data to the server through the 4G network, and the user can access the server at the PC end or the mobile phone end through the network to obtain the water quality data.
In the embodiment, the temperature acquisition module and the water temperature acquisition module both adopt a DS18B20 sensor subjected to waterproof treatment; the conductivity module adopts a DJS-1 type platinum black point conductive electrode produced by Shanghai precision scientific instruments Co., Ltd; the pH acquisition module adopts an E-201-C type pH electrode produced by Shanghai precision scientific instruments, Inc.; the dissolved oxygen acquisition module adopts a polarographic sensor; the first microcontroller module and the second microcontroller module both adopt STM32F103RBT6 chips produced by Italian semiconductor corporation; the 4G communication module uses USIM card and adopts China Mobile network.
The embodiment adopts 214 days of measured data from 1/6/2018 to 31/12/2018 to train and test, and 5136 data records are recorded in the period. Data are randomly extracted for 10 days (240 pieces) to serve as a test set, the accuracy of dissolved oxygen prediction is verified, and other data serve as a training set. As shown in FIG. 5, the predicted output value after NMEA-BP model training can approach the true value, and the overall training effect is good. The error between most dissolved oxygen values and the true value is small, and the maximum absolute error can be controlled within 0.5 mg/L.
Meanwhile, Root Mean Square Error (RMSE), Mean Relative Percentage Error (MRPE) and Mean Absolute Error (MAE) indexes are selected as judgment standards to verify the effectiveness of the method. The conventional BP model and the GA-BP model are compared with the NMEA-BP model, the weight and the threshold of the BP neural network are optimized by using the ecological Niche Mental Evolution Algorithm (NMEA), the problems of local minimum and initial parameter decision are avoided, the precision and the convergence rate of the model are improved, and therefore compared with the prediction effects of the common BP model and the GA-BP model, the NMEA-BP model has higher precision and can better predict the trend of dissolved oxygen. The interpolation evaluation index results of the three models are shown in table 1, and the comparison result in table 1 shows that when the NMEA-BP model is used for prediction, the Root Mean Square Error (RMSE), the average relative percentage error (MRPE) and the average absolute error (MAE) are lower than those of the BP model and the GA-BP model, so that the accuracy of the output value of the dissolved oxygen prediction model is higher and the dissolved oxygen prediction model is more effective.
And (3) carrying out interpolation performance evaluation comparison on the BP model, the GA-BP model and the NMEA-BP model, wherein the interpolation performance evaluation comparison is shown in the following table 1:
Figure BDA0002296154840000081
table 1.

Claims (9)

1. A method for predicting water dissolved oxygen of the Internet of things is characterized by comprising the following steps:
the prediction method comprises the following steps:
the method comprises the following steps: determining a training set and a test set;
acquiring data of air temperature, humidity, rainfall, wind speed, air pressure, water temperature, pH, conductivity and dissolved oxygen through a water quality monitoring system, dividing the data into a training set and a testing set, and setting the data of the air temperature, the humidity, the rainfall, the wind speed, the air pressure, the water temperature, the pH and the conductivity as input quantity pnThe data of dissolved oxygen is output Y;
step two: determining a BP neural network topological structure;
input quantity p according to training setnAnd the mapping corresponding relation between the output quantity Y, a three-layer topological structure is established to construct a BP neural network, the number of nodes of an input layer is n, the number of nodes of an output layer is m,
number of hidden layer nodes n1：
Figure FDA0002296154830000011
In the formula, a is an integer between 0 and 10;
the BP neural network model is as follows:
Y＝purelin(W2,1×tansig(W1,1pn+b1)+b2) (2)
wherein Y is the output, W1,1And W2,1Network connection weights from input layer to hidden layer and from hidden layer to output layer, respectively, b1And b2Network connection threshold, p, from input layer to hidden layer, hidden layer to output layer, respectivelynFor input quantity, tansig is a transfer function from an input layer to a hidden layer, and purelin is a transfer function from the hidden layer to an output layer;
the method comprises the following steps that a sigmoid function is selected as an activation function between an input layer and an output layer, and the function form is as follows:
Figure FDA0002296154830000012
step three: obtaining the optimal weight and threshold value required by the BP neural network through a niche thought evolution algorithm NMEA;
(a) group initialization
Randomly generating N individuals to form an initial population;
the score of each individual is calculated by the equations (4) and (5),
Figure FDA0002296154830000013
Figure FDA0002296154830000021
in the formula, tiAnd
Figure FDA0002296154830000022
respectively taking the actual value of the ith sample and the output value of the BP neural network model;
(b) generating a new sub-population;
sorting the individuals according to the scores of the individuals from big to small, selecting the first q individuals as winners, and forming sub-groups respectively by taking the winners as centers;
(c) generating a niche;
calculating the distance between the winners of the q sub-groups according to the formula (6), when the distance between the winners of the two sub-groups is smaller than the radius of the niche, discarding the sub-group with low score, keeping the score of the sub-group high, and DijThe distance between two sub-group winners is defined, and s is the individual coding length;
Figure FDA0002296154830000023
(d) determining optima for each niche independent search space
Reinitializing the abandoned sub-population, reselecting a winner in the niche where the abandoned sub-population is located, and turning to the step (c) to judge the niche winner score again until each niche has the winner, namely determining the optimal person of each niche independent search space;
(e) convergent operations of subgroups
Converging the sub-populations, and differentiating the populations with the lowest scores after convergence;
(f) performing diversity operation on the sub-groups to obtain optimal weight values and threshold values;
stopping iteration when the score of the optimal winner does not change any more to obtain an optimal individual, namely obtaining a group of optimal weight values and threshold values;
step four: constructing an NMEA-BP model;
substituting the optimal weight and the threshold obtained in the third step into a formula (2) to construct an NMEA-BP model which is a dissolved oxygen prediction model, wherein the optimal weight and the threshold are used as the initial weight and the threshold of the BP neural network model;
step five: and verifying the NMEA-BP model constructed in the step four according to the test set.
2. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 1, characterized in that:
the water quality monitoring system comprises an acquisition node, a sink node and a server;
the collection node comprises a first microcontroller module, a temperature collection module, a humidity collection module, an air pressure collection module, a rainfall module, an air speed module, a water temperature collection module, a conductivity module, a pH collection module, a dissolved oxygen module, a first LoRa communication module and a first power supply module;
the first microcontroller module is respectively connected with the temperature acquisition module, the humidity acquisition module, the air pressure acquisition module, the rainfall module, the air speed module, the water temperature acquisition module, the conductivity module, the pH acquisition module, the dissolved oxygen module and the first LoRa communication module;
the first power supply module is connected with other modules in the acquisition node;
the sink node comprises a second microcontroller module, a second LoRa communication module, a 4G communication module and a second power supply module;
the second microcontroller module is respectively connected with the second LoRa communication module and the 4G communication module;
the second power supply module is connected with other modules in the sink node;
the collection node sends each module data of collection to the sink node through the loRa network that first loRa communication module and second loRa communication module constitute, the sink node is through 4G network with data transmission to server.
3. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 2, characterized in that: the temperature acquisition module and the water temperature acquisition module both adopt DS18B20 sensors which are subjected to waterproof treatment.
4. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 2, characterized in that: the conductivity module adopts a DJS-1 type platinum black point conductive electrode produced by Shanghai precision scientific instruments Co.
5. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 2, characterized in that: the pH acquisition module adopts an E-201-C type pH electrode produced by Shanghai precision scientific instruments, Inc.
6. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 2, characterized in that: the dissolved oxygen acquisition module adopts a polarographic sensor.
7. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 2, characterized in that: the first microcontroller module and the second microcontroller module both adopt STM32F103RBT6 chips manufactured by Italian semiconductor corporation.
8. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 2, characterized in that: the 4G communication module uses a USIM card and adopts a China mobile network.
9. The water quality monitoring system of the internet of things water body dissolved oxygen prediction method according to claim 2, characterized in that: the user can access the server to obtain the water quality data through the network at the computer end or the mobile phone end.
CN201911202271.4A 2019-11-29 2019-11-29 Internet of things water body dissolved oxygen prediction method and water quality monitoring system Pending CN111105079A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911202271.4A CN111105079A (en) 2019-11-29 2019-11-29 Internet of things water body dissolved oxygen prediction method and water quality monitoring system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911202271.4A CN111105079A (en) 2019-11-29 2019-11-29 Internet of things water body dissolved oxygen prediction method and water quality monitoring system

Publications (1)

Publication Number Publication Date
CN111105079A true CN111105079A (en) 2020-05-05

Family

ID=70421523

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911202271.4A Pending CN111105079A (en) 2019-11-29 2019-11-29 Internet of things water body dissolved oxygen prediction method and water quality monitoring system

Country Status (1)

Country Link
CN (1) CN111105079A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112985608A (en) * 2021-02-01 2021-06-18 河北工业大学 Method and system for monitoring temperature in asphalt conveying process
CN113065687A (en) * 2021-03-16 2021-07-02 西南民族大学 Aquaculture method and system based on dissolved oxygen prediction
CN115236149A (en) * 2022-07-15 2022-10-25 中国地质调查局水文地质环境地质调查中心 Water quality detection method and system based on electrochemical sensor
CN117706045A (en) * 2024-02-06 2024-03-15 四川省德阳生态环境监测中心站 Combined control method and system for realizing atmospheric ozone monitoring equipment based on Internet of things

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112985608A (en) * 2021-02-01 2021-06-18 河北工业大学 Method and system for monitoring temperature in asphalt conveying process
CN112985608B (en) * 2021-02-01 2022-08-02 河北工业大学 Method and system for monitoring temperature in asphalt conveying process
CN113065687A (en) * 2021-03-16 2021-07-02 西南民族大学 Aquaculture method and system based on dissolved oxygen prediction
CN113065687B (en) * 2021-03-16 2023-09-05 西南民族大学 Aquaculture method and system based on dissolved oxygen prediction
CN115236149A (en) * 2022-07-15 2022-10-25 中国地质调查局水文地质环境地质调查中心 Water quality detection method and system based on electrochemical sensor
CN117706045A (en) * 2024-02-06 2024-03-15 四川省德阳生态环境监测中心站 Combined control method and system for realizing atmospheric ozone monitoring equipment based on Internet of things
CN117706045B (en) * 2024-02-06 2024-05-10 四川省德阳生态环境监测中心站 Combined control method and system for realizing atmospheric ozone monitoring equipment based on Internet of things

Similar Documents

Publication Publication Date Title
CN111105079A (en) Internet of things water body dissolved oxygen prediction method and water quality monitoring system
Chen et al. Feed intake prediction model for group fish using the MEA-BP neural network in intensive aquaculture
Hobday et al. The role of coastal ocean variation in spatial and temporal patterns in survival and size of coho salmon (Oncorhynchus kisutch)
CN107169621A (en) A kind of Dissolved Oxygen in Water Forecasting Methodology and device
CN114971295B (en) Reservoir scheduling method and system for improving habitat of birds waiting in river-passing lakes
CN115396981A (en) Intelligent monitoring system based on big data technology
CN107506857A (en) Urban lake storehouse blue-green alga bloom multi variant based on fuzzy support vector machine
Pulgar et al. Can upwelling signals be detected in intertidal fishes of different trophic levels?
Gao et al. Diurnal and seasonal patterns of methane emissions from a dairy operation in North China Plain
Knutsen et al. Genetic evidence for mixed origin of recolonized sea trout populations
Illius Linking functional responses and foraging behaviour to population dynamics
CN115828758B (en) Seawater three-dimensional prediction method and system based on improved firework algorithm optimization network
CN113033890A (en) Method for analyzing laying performance of laying hens based on vector autoregressive model
Xia et al. Environmental factor assisted chlorophyll-a prediction and water quality eutrophication grade classification: A comparative analysis of multiple hybrid models based on a SVM
Deng et al. Research of fish disease diagnosis expert system based on artificial neural networks
Ikeda et al. Metabolism and body composition of zooplankton in the cold mesopelagic zone of the southern Japan Sea
Biswas et al. Climate change impacts on fish catch in the world fishing grounds
Zhang Water Quality Substance Detection System Based on Internet of Things
CN112507533B (en) Method for constructing iris cuttlefish growth model based on marine environmental factors
Dugdale Biological modelling I
Yang et al. Real-time optimized prediction model for dissolved oxygen in crab aquaculture ponds using back propagation neural network
Hou et al. Multi-objective scheduling model of green production based on genetic algorithm under agricultural supply side structure
Shulin et al. Prediction of Soil Moisture Content Based on Improved BP Neural Network
Beseres et al. Changes in the spatial distribution of subtidal macrobenthos due to predation by white shrimp (Litopenaeus setiferus)
Geng et al. Study on prediction of dissolved oxygen content in aquaculture water

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination