CN116880602A - Temperature control method based on self-coding fuzzy neural network - Google Patents
Temperature control method based on self-coding fuzzy neural network Download PDFInfo
- Publication number
- CN116880602A CN116880602A CN202310858273.9A CN202310858273A CN116880602A CN 116880602 A CN116880602 A CN 116880602A CN 202310858273 A CN202310858273 A CN 202310858273A CN 116880602 A CN116880602 A CN 116880602A
- Authority
- CN
- China
- Prior art keywords
- self
- neural network
- output
- control
- fuzzy
- 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
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 70
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000012545 processing Methods 0.000 claims abstract description 11
- 230000009467 reduction Effects 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims description 12
- 230000004913 activation Effects 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 10
- 210000002569 neuron Anatomy 0.000 claims description 8
- 238000006243 chemical reaction Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 230000008859 change Effects 0.000 abstract description 2
- 238000009776 industrial production Methods 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 abstract description 2
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D23/00—Control of temperature
- G05D23/19—Control of temperature characterised by the use of electric means
- G05D23/20—Control of temperature characterised by the use of electric means with sensing elements having variation of electric or magnetic properties with change of temperature
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a temperature control method based on a self-coding fuzzy neural network, and belongs to the technical field of temperature control. The temperature control method comprises the following steps: collecting temperature sensor data and preprocessing; learning and dimension reduction processing are carried out on the preprocessed temperature data by using a self-coding neural network, and a low-dimension representation is generated; controlling the low-dimensional representation and the control error by using a fuzzy controller, and generating a corresponding control signal; and feeding back a control signal to the temperature control system for real-time control. The invention uses the self-coding neural network to generate the low-dimensional representation by the temperature data dimension reduction processing, uses the fuzzy controller to control the low-dimensional representation, and realizes the high efficiency and the precision of the temperature control by controlling the temperature change. Meanwhile, the method has the advantages of strong adaptability, good robustness and the like, and can be suitable for various complex industrial production and manufacturing environments.
Description
Technical Field
The invention belongs to the technical field of temperature control, and particularly relates to a temperature control method based on a self-coding fuzzy neural network.
Background
Current temperature control systems generally use linear control methods such as PID, however, these conventional methods have some limitations in coping with nonlinear systems. Conventional PID control methods often rely on accurate models and linear assumptions, which are difficult to effectively address complex nonlinear temperature control problems. Furthermore, conventional control methods typically require manual adjustment of parameters, which are difficult to optimize in real time for dynamically changing systems. Therefore, in the field of temperature control, there is a need to find a control method that is more adaptive and self-adaptive.
In recent years, self-encoding neural networks in the field of deep learning have been widely studied and applied for nonlinear modeling and control tasks. However, the existing self-coding fuzzy neural network method still has some limitations in terms of temperature control. For example, they may be limited by problems with the particular distribution of input data, the selection of fuzzy aggregation rules, the design of the network architecture, etc. In order to overcome the limitations of the existing self-coding fuzzy neural network and improve the accuracy and the robustness of temperature control, the invention provides a temperature control system based on the self-coding fuzzy neural network.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a temperature control method based on a self-coding fuzzy neural network. The invention combines the self-coding neural network with the fuzzy control method, introduces a fuzzy aggregation rule and an improved network structure, so that the network can better cope with the problems of ambiguity and uncertainty, and improves the performance of the temperature control system in a complex environment.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows: a temperature control method based on a self-coding fuzzy neural network comprises the following steps:
(1) Acquiring temperature data to be controlled, and preprocessing the temperature data to obtain preprocessed data;
(2) The preprocessed data is subjected to learning and dimension reduction processing by using a self-coding neural network, and low-dimension representation is generated;
(3) Using a fuzzy controller to take the low-dimensional representation and the control error as input, and then inverting the output control increment to obtain a control signal;
(4) And feeding back a control signal to the temperature control system for real-time control.
Compared with other neural networks, the self-coding neural network can be trained without labeling data, and is therefore suitable for a scene of a large-scale data set and label-free data. The self-coding neural network has the advantages in the tasks of data preprocessing, feature learning, data dimension reduction and the like; important features of the data can be extracted by learning a low-dimensional representation of the input data, which is helpful for reducing the data dimension and retaining key information; compared with the traditional linear model, the self-coding neural network has stronger nonlinear modeling capability, can learn complex nonlinear relations through the combination of multi-layer neurons and activation functions, and is suitable for processing the highly nonlinear temperature control problem.
The self-coding neural network adopts a multi-layer perceptron architecture, and the expression form is as follows:
where x is input data; w (W) 1 And b 1 The weight and bias of the self-coding neural network are respectively; f is the Logsig activation function, f (W 1 x+b 1 ) Expressed as a delivery pattern of the self-encoding neural network in a previous layer;is expressed as the final output form of the neural network; w (W) 2 And b 2 Respectively the weight and the bias of the fuzzy controller;
the loss function of the self-encoding neural network is defined as:
L(θ)＝L AE (θ)+L FC (θ)
wherein L is AE (θ) is the reconstruction error, L, of the improved self-encoding neural network FC And (θ) is the error of the fuzzy controller, and θ is a parameter of the improved self-encoding neural network.
The low-dimensional representation of the present invention is represented as representing data using fewer features or dimensions relative to the original data.
As a preferred embodiment of the present invention, the error L of the fuzzy controller FC The expression form of (θ) is:
where N is the number of samples, i is the ith sample, y i For the input of the fuzzy controller,is the desired output.
As a preferred embodiment of the present invention, the hidden layer and the output layer of the self-encoding neural network both contain a log sig activation function defined asx is the input data.
In the forward propagation of the self-encoding neural network, the output of each neuron can be calculated by multiplying the input vector x with the weight matrix W and adding the bias vector b by the activation function f (x), the specific propagation is as follows:
where x is the input of the neuron, y is the output of the neuron, W is the weight of the neuron, b is the bias, b 1 Is the bias of the self-encoding neural network.
As a preferred embodiment of the present invention, in the step (2), the preprocessed temperature data is input into the self-coding neural network, training is performed through a back propagation algorithm, features of the data are learned, and the preprocessed temperature data is compressed to a low-dimensional representation, which specifically includes the following steps:
s1: for a temperature dataset { x } containing N samples 1 ,x 2 ,...,x N Each sample x i ＝{x i1 ,x i2 ,...,x im M features, importing the data set into a self-encoding neural network;
s2: performing dimension reduction processing on the data by using a self-encoder in the self-encoding neural network, and repeatedly and iteratively solving to minimize a reconstruction error, wherein:
let the node number of hidden layer z be k l Wherein the input layer and the output layer have n l The goal of the self-encoder is to minimize the error between the input and output, i.e. to minimize the reconstruction error:
wherein N represents the number of samples, L (x i ,f(g(x i ) X) is input x) i And output f (g (x) i ) Mean square error between g (x) i ) Representing the conversion of the input layer into the hidden layer, f (g (x i ) A) represents a transition from the hidden layer to the output layer;is a sparse regularization term used for controlling the activation number of the neurons of the self-encoding neural network; θ is the self-encoder model parameter; l is the number of hidden layers, i is the i-th node of the currently computed hidden layer, j is the input of the current computationThe j-th node of the layer and the output layer;
s3: training is completed when convergence conditions are reached to set the required minimum reconstruction error, using the output from the hidden layer of the encoder as a reduced-dimension representation of the dataset:
x i →g(x i )→z i
wherein g (x) i ) Is the conversion of the input layer to the hidden layer, z i Is the output of the hidden layer, i.e., the dimension-reduced representation;
the convergence condition is that the reconstruction error is less than 1e-6.
In the step (3), the reduced-dimension representation and the control error are decomposed into a plurality of local models, a fuzzy controller is designed on each local model, and then the output of each fuzzy controller is synthesized by adopting a weighted average method to obtain the control increment.
As a preferred embodiment of the present invention, the fuzzy controller is a T-S fuzzy controller.
The input of the T-S fuzzy controller comprises a dimension reduction representation and a control error, and the output is a control increment, which can be represented as:
wherein Δh a Is the control increment on the a-th local model, w a Is the weight of the a-th local model.
The invention can be inverted by a PID controller, expressed as:
where u (t) is the output of the PID controller and e (t) is the control error.
The PID controller total control increment deltau is therefore expressed as:
because the total control increment Deltau of the PID controller is the output result of the T-S fuzzy controller, the control error needs to be recalculated according to the actual situation, and the specific calculation formula is as follows:
e(t)＝y ref (t)-y(t)
in summary, the output control signal expression when the T-S fuzzy controller integrates the PID controller is:
in the step (3), inversion is performed by using a PID controller, and the expression of the output control signal when the T-S fuzzy controller integrates the PID controller is as follows:
where t is the current time, D h a Is the control increment, w, on the a-th local model of the T-S fuzzy controller a Is the weight of the a local model, K p 、K i 、K d Proportional, integral, differential coefficients, respectively; y is ref (t) is a set reference output value, y (t) is an actual output value, N is the number of samples, and a is the a-th local model.
In the step (1), the preprocessing is to normalize the collected temperature data, map the value ranges of all the input parameter raw data to [0,1], and the normalization is as follows:
wherein x is new Is the normalized parameter value, x is the parameter value before normalization, x min Is the minimum parameter value, x, within the sample max Is the maximum parameter value within the sample.
As a preferred embodiment of the present invention, in the step (1), the collected data is divided into a test set and a training set before the preprocessing.
Compared with the prior art, the invention has the beneficial effects that:
(1) The invention uses the self-coding neural network to generate the low-dimensional representation by the temperature data dimension reduction processing, uses the fuzzy controller to control the low-dimensional representation and the control error, realizes the high efficiency and the precision of the temperature control by controlling the temperature change, and can adapt to various complex industrial production and manufacturing environments.
(2) The method combining the self-coding neural network and the fuzzy control ensures that the self-coding fuzzy neural network can be better suitable for a nonlinear temperature control system, automatically extracts the characteristics and rules of the system from data by using a data-driven modeling method, can be suitable for different working conditions and system changes, and finally can better predict the system behavior and optimize the control strategy, thereby realizing better control effect, providing technical reference for modeling and control of the nonlinear system in industry, overcoming the limitation of the prior art and showing better performance in the nonlinear temperature control system.
Drawings
Fig. 1 is a network structure diagram of the self-encoding neural network according to the present invention.
Fig. 2 is a control flow chart of the temperature control method based on the self-coding fuzzy neural network.
Fig. 3 is a graph comparing the final predicted control result and the actual result of example 1 of the present invention.
Fig. 4 is a graph of the predicted mean square error of the temperature control method based on the self-coding fuzzy neural network according to the embodiment 1 of the present invention compared with the actual temperature.
FIG. 5 is a graph showing the response of the temperature control method based on the self-encoding fuzzy neural network and the conventional PID control method in the present invention.
Detailed Description
For a better description of the objects, technical solutions and advantages of the present invention, the present invention will be further described with reference to the following specific examples.
Example 1
A temperature control method based on a self-coding fuzzy neural network comprises the following steps:
(1) The temperature data to be controlled are collected through the temperature collection module, the temperature data are divided into a test set and a training set, normalization processing is carried out on the temperature data, the value ranges of all the input parameter original data are mapped to [0,1], and the normalization processing formula is as follows:
wherein x is new Is the normalized parameter value, x is the parameter value before normalization, x min Is the minimum parameter value, x, within the sample max Is the maximum parameter value within the sample.
(2) As shown in fig. 1, the data preprocessed in the training set is trained by using a self-coding neural network through a back propagation algorithm, features of the data are learned, and the preprocessed temperature data is compressed to a low-dimensional representation for setting control rules of a subsequent fuzzy controller, which specifically includes the following steps:
s1: for a dataset { x } containing N samples 1 ,x 2 ,...,x N Each sample x i ＝{x i1 ,x i2 ,...,x im M features;
s2: performing dimension reduction processing on the data by using a self-encoder:
let the number of nodes of the hidden layer z be k, where both the input layer and the output layer have n l The goal of the self-encoder is to minimize the error between the input and output, i.e. to minimize the reconstruction error:
wherein L (x) i ,f(g(x i ) X) is input x) i And output f (g (x) i ) Mean square error between g (x) i ) Representing the conversion of the input layer into the hidden layer, f (g (x i ) A) represents a transition from the hidden layer to the output layer;is a sparse regularization term used for controlling the activation number of the neurons of the self-encoding neural network; θ is the self-encoder model parameter; l is the number of hidden layers, i is the i-th node of the currently calculated hidden layer, and j is the j-th node of the currently calculated input layer and output layer.
S3: training is completed when a convergence condition is reached that sets the required minimum reconstruction error, i.e. the convergence condition is less than 1e-6, using the output of the hidden layer of the self-encoder as a reduced dimension representation of the dataset:
x i →g(x i )→z i
wherein g (x) i ) Is the conversion of the input layer to the hidden layer, z i Is the output of the hidden layer, i.e. the reduced-dimension representation.
The self-coding neural network adopts a multi-layer perceptron architecture, and the expression form is as follows:
where x is input data; w (W) 1 And b 1 The weight and bias of the self-coding neural network are respectively; f is the Logsig activation function, f (W 1 x+b 1 ) Expressed as a delivery pattern of the self-encoding neural network in a previous layer;is expressed as the final output form of the neural network; w (W) 2 And b 2 Respectively the weight and the bias of the fuzzy controller;
the loss function of the self-encoding neural network is defined as:
L(θ)＝L AE (θ)+L FC (θ)
wherein L is AE (θ) is the reconstruction error of the self-encoding neural network, L FC And (θ) is an error of the fuzzy controller, and θ is a parameter of the self-encoding neural network.
(3) First, a fuzzy controller is designed: constructing a fuzzy controller by using a T-S fuzzy model, taking low-dimensional representation and control errors as inputs, setting an input membership function of the fuzzy controller as a Gaussian membership function, training the fuzzy controller, and constructing an optimal fuzzy control rule; decomposing the low-dimensional representation and control error into a plurality of local models, designing a fuzzy controller on each local model, and synthesizing the output of each fuzzy controller by adopting a weighted average method to obtain a control increment;
then, inversion is carried out on the output control increment through a PID controller to obtain a control signal, and an output control signal expression is obtained when the T-S fuzzy controller integrates the PID controller:
wherein Δh a Is the control increment, w, on the a-th local model of the T-S fuzzy controller a Is the weight of the a local model, K p 、K i 、K d Respectively proportional, integral and differential coefficients, in this example the original PID parameters are set to K p ＝0.001，K i ＝0.2，K d ＝0.0008；y ref (t) is a set reference output value, y (t) is an actual output value, N is the number of samples, and a is the a-th local model.
(4) And feeding back a control signal to the temperature control system for real-time control. And feeding back an output signal of the fuzzy controller to a temperature control system for real-time control.
As can be seen from fig. 3 and fig. 4, the temperature control method for a self-coding fuzzy neural network for a nonlinear system according to the present invention can well predict the temperature of a nonlinear model, and the image of fig. 5 shows that the control method for a self-coding fuzzy neural network according to the present invention (i.e., the temperature control method for a self-coding fuzzy neural network in the figure) has smaller fluctuation and can reach the expected temperature value more stably compared with the conventional PID control scheme (i.e., the PID in the figure), and meanwhile, compared with the overshoot of 0.88% in the conventional PID, the overshoot of the temperature control method for a self-coding fuzzy neural network for a nonlinear system according to the present invention is only 0.07%, and due to the limitation of the maximum temperature in the high temperature reaction, the smaller overshoot can effectively avoid the safety risk caused by uncontrollability of high temperature.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the scope of the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted equally without departing from the spirit and scope of the technical solution of the present invention.
Claims (10)
1. The temperature control method based on the self-coding fuzzy neural network is characterized by comprising the following steps of:
(1) Acquiring temperature data to be controlled, and preprocessing the temperature data to obtain preprocessed data;
(2) The preprocessed data is subjected to learning and dimension reduction processing by using a self-coding neural network, and low-dimension representation is generated;
(3) Using a fuzzy controller to take the low-dimensional representation and the control error as input, and then inverting the output control increment to obtain a control signal;
(4) Feeding back a control signal to a temperature control system for real-time control;
the self-coding neural network adopts a multi-layer perceptron architecture, and the expression form is as follows:
where x is input data; w (W) 1 And b 1 The weight and bias of the self-coding neural network are respectively; f is the Logsig activation function, f (W 1 x+b 1 ) Expressed as a delivery pattern of the self-encoding neural network in a previous layer;is expressed as the final output form of the neural network; w (W) 2 And b 2 Respectively the weight and the bias of the fuzzy controller;
the loss function of the self-encoding neural network is defined as:
L(θ)＝L AE (θ)+L FC (θ)
wherein L is AE (θ) is the reconstruction error of the self-encoding neural network, L FC And (θ) is an error of the fuzzy controller, and θ is a parameter of the self-encoding neural network.
2. The temperature control method based on the self-coding fuzzy neural network as claimed in claim 1, wherein the specific calculation formula of the control error is as follows:
e(t)＝y ref (t)-y(t)
wherein y is ref (t) is a set reference output value, y (t) is an actual output value, and e (t) is a control error.
3. The method for controlling temperature based on self-encoding fuzzy neural network as claimed in claim 1, wherein the error L of the fuzzy controller FC The expression form of (θ) is:
where N is the number of samples, i is the ith sample, y i In order for the output to be desired,is the input quantity of the self-coding neural network.
4. The method of claim 1, wherein each layer of the self-encoding fuzzy neural network comprises a Logsig activation function defined asx is input data.
5. The method for controlling temperature based on the self-encoding fuzzy neural network according to claim 1, wherein in the step (2), the preprocessed temperature data is input into the self-encoding neural network, training is performed through a back propagation algorithm, features of the data are learned, and the preprocessed temperature data is compressed to a low-dimensional representation, and the method specifically comprises the following steps:
s1: for a dataset { x } containing N samples 1 ,x 2 ,…,x N Each sample x i ＝{x i1 ,x i2 ,…,x im M features;
s2: performing dimension reduction processing on the data by using a self-encoder:
let the node number of hidden layer z be k l Wherein the input layer and the output layer have n l The goal of the self-encoder is to minimize the error between the input and output, i.e. to minimize the reconstruction error:
wherein L (x) i ,f(g(x i ) X) is input x) i And output f (g (x) i ) Mean square error between g (x) i ) Representing the conversion of the input layer into the hidden layer, f (g (x i ) A) represents a transition from the hidden layer to the output layer;is a sparse regularization term used for controlling the activation number of the neurons of the self-encoding neural network; θ isSelf-encoder model parameters; l is the number of hidden layers, i is the i-th node of the currently calculated hidden layer, and j is the j-th node of the currently calculated input layer and output layer.
S3: training is completed when convergence conditions are reached to set the required minimum reconstruction error, using the output from the hidden layer of the encoder as a reduced-dimension representation of the dataset:
x i →g(x i )→z i
wherein g (x) i ) Is the conversion of the input layer to the hidden layer, z i Is the output of the hidden layer, i.e. the reduced-dimension representation.
6. The method of claim 1, wherein in the step (3), the reduced-dimension representation and the control error are decomposed into a plurality of local models, a fuzzy controller is designed on each local model, and then the output of each fuzzy controller is synthesized by adopting a weighted average method to obtain the control increment.
7. The method of claim 6, wherein the fuzzy controller is a T-S fuzzy controller.
8. The method of temperature control based on a self-encoding fuzzy neural network according to claim 7, wherein in the step (3), inversion is performed by using a PID controller, and the T-S fuzzy controller integrates an output control signal expression of the PID controller:
wherein Dh is a Is the control increment, w, on the a-th local model of the T-S fuzzy controller a Is the weight of the a local model, K p 、K i 、K d Proportional, integral, differential coefficients, respectively; y is ref (t) is the set reference output value, y (t) is the actual outputAnd (3) outputting a value, wherein N is the number of samples, and a is the a-th local model.
9. The method of claim 1, wherein in the step (1), the preprocessing is to normalize the collected temperature data, and the value ranges of all the input parameter raw data are mapped to [0,1], and the normalization formula is:
wherein x is new Is the normalized parameter value, x is the parameter value before normalization, x min Is the minimum parameter value, x, within the sample max Is the maximum parameter value within the sample.
10. The method of claim 1, wherein in the step (1), the collected data is divided into a test set and a training set before preprocessing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310858273.9A CN116880602A (en) | 2023-07-13 | 2023-07-13 | Temperature control method based on self-coding fuzzy neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310858273.9A CN116880602A (en) | 2023-07-13 | 2023-07-13 | Temperature control method based on self-coding fuzzy neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116880602A true CN116880602A (en) | 2023-10-13 |
Family
ID=88254364
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310858273.9A Pending CN116880602A (en) | 2023-07-13 | 2023-07-13 | Temperature control method based on self-coding fuzzy neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116880602A (en) |
-
2023
- 2023-07-13 CN CN202310858273.9A patent/CN116880602A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Remaining useful life prediction using a novel feature-attention-based end-to-end approach | |
| Abiyev | Fuzzy wavelet neural network based on fuzzy clustering and gradient techniques for time series prediction | |
| CN114678080B (en) | Converter end point phosphorus content prediction model, construction method and phosphorus content prediction method | |
| CN113177633A (en) | Deep decoupling time sequence prediction method | |
| CN114839881B (en) | Intelligent garbage cleaning and environmental parameter big data Internet of things system | |
| CN115016276B (en) | Intelligent water content adjustment and environment parameter Internet of things big data system | |
| CN111475546A (en) | Financial time sequence prediction method for generating confrontation network based on double-stage attention mechanism | |
| CN114022311A (en) | Comprehensive energy system data compensation method for generating countermeasure network based on time sequence condition | |
| CN114444660A (en) | Short-term power load prediction method based on attention mechanism and LSTM | |
| CN115983493A (en) | Water quality prediction method based on multitask learning and deep learning model | |
| CN113780664A (en) | Time sequence prediction method based on TDT-SSA-BP | |
| CN114403486A (en) | Intelligent control method of airflow type cut-tobacco drier based on local peak value coding circulation network | |
| CN117574776A (en) | Task planning-oriented model self-learning optimization method | |
| Kim | Parameter tuning of fuzzy neural networks by immune algorithm | |
| CN113283642A (en) | Poultry feed detection and formula system | |
| CN113128666A (en) | Mo-S-LSTMs model-based time series multi-step prediction method | |
| CN113255739A (en) | Fish feed detection and formula system | |
| CN111863153A (en) | Method for predicting total amount of suspended solids in wastewater based on data mining | |
| CN116880602A (en) | Temperature control method based on self-coding fuzzy neural network | |
| CN115062764B (en) | Intelligent illuminance adjustment and environmental parameter Internet of things big data system | |
| CN114970745B (en) | Intelligent security and environment big data system of Internet of things | |
| CN115619563A (en) | Stock price analysis method based on neural network | |
| Ghazali et al. | Dynamic ridge polynomial neural networks in exchange rates time series forecasting | |
| Liu et al. | STA-APSNFIS: STA-optimized adaptive pre-sparse neuro-fuzzy inference system for online soft sensor modeling | |
| CN114372181A (en) | Intelligent planning method for equipment production based on multi-mode data |
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 |