Method, system and apparatus for intelligently monitoring aquafarm with multi-dimensional panoramic perception

A system for intelligently monitoring an aquafarm with multi-dimensional panoramic perception, comprising:

• a perception information acquisition module configured to acquire multi-dimensional panoramic perception information of an aquafarm with sensors, wherein the multi-dimensional panoramic perception information comprises water quality information, atmospheric environmental information, and electrical device state information in the aquafarm;
• the water quality information comprises an oxygen content, a pH value, a conductivity, a temperature, and a turbidity;
• the atmospheric environmental information comprises a temperature, a humidity, an illumination intensity, and an atmospheric pressure intensity in an atmospheric environment;
• and the electrical device state information in the aquafarm comprises currents, voltages, surface temperatures, and electric powers of electrical devices;
• a preprocessing module configured to preprocess the multi-dimensional panoramic perception information with statistical analysis to obtain preprocessed multi-dimensional panoramic perception information;
• a feature extraction module configured to perform feature extraction on the preprocessed multi-dimensional panoramic perception information to determine a threshold range of the preprocessed multi-dimensional panoramic perception information;
• a time series analysis module configured to analyze the preprocessed multi-dimensional panoramic perception information with time series analysis to obtain a linear fitting model;
• a water quality forecasting-alarming model construction module configured to construct a water quality forecasting-alarming model according to the linear fitting model;
• a data fusion module configured to fuse preprocessed water quality information and preprocessed electrical device state information to obtain fused data;
• a training module configured to input the fused data to a convolutional neural network (CNN) for training to obtain an electrical device fault diagnosis model;
• and
• a monitoring module configured to monitor the aquafarm with the water quality forecasting-alarming model and the electrical device fault diagnosis model, and give an alarm in response to a substandard water quality or a fault in each of the electrical devices;
• wherein the water quality forecasting-alarming model construction module comprises:
• a clustering unit configured to cluster the preprocessed multi-dimensional panoramic perception information with a fuzzy clustering algorithm according to the linear fitting model to obtain cluster centers;
• a model construction unit configured to take the preprocessed multi-dimensional panoramic perception information as a sample set to construct a Takagi-Sugeno (T-S) fuzzy model;
• and
• a parameter identification unit configured to identify parameters of the T-S fuzzy model with the cluster centers to obtain the water quality forecasting-alarming model;
• and
• a target function of the fuzzy clustering algorithm is given by:
• J
• m
• (
• U
• ,
• V
• )
• =
• ∑
• k
• =
• 1
• N
• ∑
• i
• =
• 1
• c
• (
• u
• ik
• )
• m
• ⁢
• 
• X
• k
• -
• υ
• i
• 
• 2
• +
• α
• ⁢
• ∑
• i
• =
• 1
• c
• 
• υ
• i
• (
• 0
• )
• -
• υ
• i
• 
• 2
• ,
• wherein, Xk is a kth sample, N is a number of samples, uik is a degree of membership of the sample Xk to a cluster center vi, m represents a number of clusters, vi represents an ith cluster center, c represents a number of the cluster centers, α is a regularization parameter, vi(0) represents a priori estimated value of vi, V is a cluster center matrix, V=(v1, v2, . . . , vc), and U is a fuzzification matrix.

An apparatus for intelligently monitoring an aquafarm with multi-dimensional panoramic perception, comprising:

• a multi-parameter water quality sensor module configured to acquire water quality information of an aquafarm;
• a meteorological sensor module configured to acquire atmospheric environmental information;
• an aquafarm device monitoring module configured to acquire electrical device state information in the aquafarm, the water quality information, the atmospheric environmental information and the electrical device state information in a fishing ground forming multi-dimensional state panoramic perception information;
• an intelligent gateway in communication connection with the multi-parameter water quality sensor module, the meteorological sensor module, and the aquafarm device monitoring module;
• and
• an edge computing terminal provided thereon with a high-speed power line carrier (HPLC) communication router and a long term evolution (LTE) communication router, wherein both the HPLC communication router and the LTE communication router are in communication connection with the intelligent gateway;
• and the edge computing terminal is configured to monitor the aquafarm according to the water quality information, the atmospheric environmental information and the electrical device state information, and give an alarm in response to a substandard water quality or a fault in each of electrical devices;
• wherein the edge computing terminal is specifically configured to:
• preprocess the multi-dimensional panoramic perception information with statistical analysis to obtain preprocessed multi-dimensional panoramic perception information;
• perform feature extraction on the preprocessed multi-dimensional panoramic perception information to determine a threshold range of the preprocessed multi-dimensional panoramic perception information;
• analyze the preprocessed multi-dimensional panoramic perception information with time series analysis to obtain a linear fitting model;
• construct a water quality forecasting-alarming model according to the linear fitting model;
• fuse preprocessed water quality information and preprocessed electrical device state information to obtain fused data;
• input the fused data to a convolutional neural network (CNN) for training to obtain an electrical device fault diagnosis model;
• and
• monitor the aquafarm with the water quality forecasting-alarming model and the electrical device fault diagnosis model, and give the alarm in response to the substandard water quality or the fault in each of the electrical devices;
• constructing the water quality forecasting-alarming model according to the linear fitting model comprises:
• clustering the preprocessed multi-dimensional panoramic perception information with a fuzzy clustering algorithm according to the linear fitting model to obtain cluster centers;
• taking the preprocessed multi-dimensional panoramic perception information as a sample set to construct a Takagi-Sugeno (T-S) fuzzy model;
• and
• identifying parameters of the T-S fuzzy model with the cluster centers to obtain the water quality forecasting-alarming model;
• and
• a target function of the fuzzy clustering algorithm is given by:
• J
• m
• (
• U
• ,
• V
• )
• =
• ∑
• k
• =
• 1
• N
• ∑
• i
• =
• 1
• c
• (
• u
• ik
• )
• m
• ⁢
• 
• X
• k
• -
• υ
• i
• 
• 2
• +
• α
• ⁢
• ∑
• i
• =
• 1
• c
• 
• υ
• i
• (
• 0
• )
• -
• υ
• i
• 
• 2
• ,
• wherein, Xk is a kth sample, N is a number of samples, uik is a degree of membership of the sample Xk to a cluster center vi, m represents a number of clusters, vi represents an ith cluster center, c represents a number of the cluster centers, α is a regularization parameter, vi(0) represents a priori estimated value of vi, V is a cluster center matrix, V=(v1, v2, . . . , vc), and U is a fuzzification matrix.