CN117476125A - Dried beancurd stick raffinate recovery data processing system based on data analysis - Google Patents
Dried beancurd stick raffinate recovery data processing system based on data analysis Download PDFInfo
- Publication number
- CN117476125A CN117476125A CN202311812985.3A CN202311812985A CN117476125A CN 117476125 A CN117476125 A CN 117476125A CN 202311812985 A CN202311812985 A CN 202311812985A CN 117476125 A CN117476125 A CN 117476125A
- Authority
- CN
- China
- Prior art keywords
- model
- dissolved oxygen
- data
- power
- temperature value
- 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.)
- Granted
Links
- 235000013527 bean curd Nutrition 0.000 title claims abstract description 50
- 238000012545 processing Methods 0.000 title claims abstract description 33
- 238000007405 data analysis Methods 0.000 title claims abstract description 28
- 238000011084 recovery Methods 0.000 title claims abstract description 20
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims abstract description 120
- 229910052760 oxygen Inorganic materials 0.000 claims abstract description 120
- 239000001301 oxygen Substances 0.000 claims abstract description 120
- 238000012549 training Methods 0.000 claims abstract description 90
- 239000000446 fuel Substances 0.000 claims abstract description 57
- 230000000813 microbial effect Effects 0.000 claims abstract description 57
- 238000010248 power generation Methods 0.000 claims abstract description 50
- 238000000034 method Methods 0.000 claims abstract description 36
- 238000006243 chemical reaction Methods 0.000 claims abstract description 30
- 238000013256 Gubra-Amylin NASH model Methods 0.000 claims abstract description 25
- 239000007788 liquid Substances 0.000 claims abstract description 25
- 238000004064 recycling Methods 0.000 claims abstract description 10
- 239000000523 sample Substances 0.000 claims description 60
- 230000008569 process Effects 0.000 claims description 29
- 238000003062 neural network model Methods 0.000 claims description 21
- 239000002351 wastewater Substances 0.000 claims description 15
- 238000013528 artificial neural network Methods 0.000 claims description 14
- 230000003993 interaction Effects 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 8
- 238000013459 approach Methods 0.000 claims description 7
- 239000002699 waste material Substances 0.000 claims description 6
- 239000000758 substrate Substances 0.000 claims description 5
- 239000012295 chemical reaction liquid Substances 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 230000010365 information processing Effects 0.000 claims description 3
- 239000013074 reference sample Substances 0.000 claims description 3
- 230000006870 function Effects 0.000 description 45
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 6
- 230000008901 benefit Effects 0.000 description 4
- 238000009826 distribution Methods 0.000 description 4
- 244000005700 microbiome Species 0.000 description 4
- 239000010865 sewage Substances 0.000 description 4
- 230000001276 controlling effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000001105 regulatory effect Effects 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 2
- 235000010469 Glycine max Nutrition 0.000 description 2
- 244000068988 Glycine max Species 0.000 description 2
- 239000000370 acceptor Substances 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 229910052799 carbon Inorganic materials 0.000 description 2
- 239000013043 chemical agent Substances 0.000 description 2
- 238000004134 energy conservation Methods 0.000 description 2
- 239000007789 gas Substances 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000006722 reduction reaction Methods 0.000 description 2
- 239000013049 sediment Substances 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 238000004065 wastewater treatment Methods 0.000 description 2
- 241000894006 Bacteria Species 0.000 description 1
- MYMOFIZGZYHOMD-UHFFFAOYSA-N Dioxygen Chemical compound O=O MYMOFIZGZYHOMD-UHFFFAOYSA-N 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000005273 aeration Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 235000019577 caloric intake Nutrition 0.000 description 1
- 150000001720 carbohydrates Chemical class 0.000 description 1
- 230000003197 catalytic effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 229910001882 dioxygen Inorganic materials 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000003411 electrode reaction Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000004060 metabolic process Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000010815 organic waste Substances 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000010992 reflux Methods 0.000 description 1
- 230000027756 respiratory electron transport chain Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/10—Analysis or design of chemical reactions, syntheses or processes
-
- 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
- G06N3/0475—Generative 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
- G06N3/094—Adversarial learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/30—Hydrogen technology
- Y02E60/50—Fuel cells
Abstract
The invention discloses a dried beancurd stick residual liquid recycling data processing system based on data analysis, which relates to the field of power management, and the technical scheme is characterized by comprising the following steps: the method comprises the steps of constructing a microbial fuel cell MFC reaction model, wherein a data acquisition unit is used for acquiring temperature values and dissolved oxygen quantities of corresponding time sequences in current period time, acquiring power generation power of each temperature value and corresponding time sequence of the dissolved oxygen quantity, a sample expansion unit expands a sample data table through a GAN model, generating a pseudo data table, merging the sample data table and the pseudo data table to form a new training set, and a model training unit carries out training according to the training set to acquire an expression relation model of the temperature values, the dissolved oxygen quantities and the power generation power; the judging unit is used for analyzing and judging the expected power and the real-time power, and the adjusting unit is used for adjusting the temperature value and the dissolved oxygen amount in the MFC based on the PID control model. Realizing the rapid and effective treatment of data in the recovery process of the dried beancurd stick residual liquid.
Description
Technical Field
The invention relates to the field of dried beancurd stick residual liquid recovery management, in particular to a dried beancurd stick residual liquid recovery data processing system based on data analysis.
Background
The dried beancurd stick wastewater is residual liquid left after soybean pulp is ground, boiled, coagulated and dried to form dried beancurd sticks. But the soybean utilization rate is low in the production process of the dried beancurd sticks, only about 50 percent of protein, about 50 percent of fat and about 20 percent of carbohydrate are utilized, so that the dried beancurd sticks waste water contains a large amount of organic matters, and the dried beancurd sticks waste water is directly discharged into the environment, so that precious resources are wasted, the ecological environment is destroyed, and therefore, the recycling of dried beancurd sticks remains is an important economic requirement, at present, the microbial fuel cell (Microbial fuel cell, MFC) is rapidly developed, and the technology integrates a microbial power generation technology and a sewage treatment technology, and can treat organic waste water and obtain electric energy. Recently, technologies for microbial power generation have been rapidly developed, MFC is increasingly widely used, and Microbial Fuel Cells (MFC) are regarded as a novel energy conversion device with dual functions of capacity and sewage treatment, and are attracting attention due to their wide application fields. The microbial fuel cell converts chemical energy into electric energy by utilizing the catalytic activity of electricity-generating microorganisms, acquires electric energy from organic matters in sewage, has the advantages of energy conservation, emission reduction, lower cost and the like, has good effect on the aspect of treating organic sewage, and can be used as a matrix of MFC (micro-organic fuel cell) due to the fact that high-concentration organic wastewater is contained, so that the microbial fuel cell can be applied to recycling of dried beancurd stick wastewater.
In the process of treating the dried beancurd stick wastewater by the microbial fuel cell, main parameters affecting the performance of the microbial fuel cell comprise an electrode, temperature, dissolved oxygen concentration, a substrate and the like, wherein the temperature and the oxygen concentration are factors which are externally influencing the environment and are easiest to control and have higher influence on efficiency, so that the reaction temperature and the oxygen inlet of the microbial fuel cell MFC are control variables with lower cost and higher economic benefit by controlling the reaction temperature and the oxygen inlet of the microbial fuel cell MFC, but fluctuation exists in the data in the reaction process of the microbial fuel cell, so that the output data has fluctuation, the data is more complex, the control difficulty of the temperature and the oxygen inlet amount is further increased, and therefore, the data generated in the control process need to be subjected to relevant treatment, the effective analysis of the data is realized, and the control effect is enhanced.
Disclosure of Invention
Aiming at the problems of high data analysis difficulty, complex battery reaction and difficult data relation expression in the power management process in the prior art, the invention aims to provide a dried beancurd stick residual liquid recovery data processing system based on data analysis, which realizes rapid analysis and processing of data in the dried beancurd stick residual liquid recovery process.
In order to achieve the above purpose, the present invention provides the following technical solutions:
a data analysis-based beancurd stick raffinate recovery data processing system, the data analysis-based beancurd stick raffinate recovery data processing system comprising:
the reaction module is used for constructing an MFC reaction model of the microbial fuel cell, constructing a double-chamber MFC by taking waste liquid of dried beancurd sticks as a substrate, setting SA-PQ-11/CF as an anode, taking dried beancurd sticks waste water as an anolyte and CF as a cathode, and constructing an MFC power generation system;
the sample model building module comprises a data acquisition unit, a sample expansion unit and a model training unit, wherein the data acquisition unit is used for acquiring temperature values and dissolved oxygen amounts of corresponding time sequences in the current period time, acquiring power generation power of each temperature value and corresponding time sequence of the dissolved oxygen amounts, forming a sample data table, the sample expansion unit expands the sample data table through a GAN model, generating a pseudo data table, merging the sample data table and the pseudo data table to form a new training set, and the model training unit performs training according to the training set to acquire an expression relation model of the temperature values, the dissolved oxygen amounts and the power generation power;
the regulation and control module comprises a judging unit and an adjusting unit, wherein the judging unit is used for analyzing and judging expected power generation and real-time power generation, judging whether to start the adjusting unit, and the adjusting unit adjusts the temperature value and the dissolved oxygen in the microbial fuel cell MFC based on the PID control model.
Preferably, the power generation output of the microbial fuel cell MFC is measured by a power sampling device, the temperature of the reaction liquid is collected and recorded by a temperature sensor, and the dissolved oxygen of the liquid is measured and recorded by a dissolved oxygen analyzer.
Preferably, in the sample expansion unit, the training process of expanding the sample data table by the GAN model and outputting the reference sample is as follows:
constructing a framework of a GAN model: selection generatorDiscriminator->Neural network structure and random noise of (a)Setting the iteration times;
training a GAN model according to the objective function to obtain a trained GAN model, inputting each group of data of the sample data table into the trained GAN model, and setting and generating Q groups of pseudo data tables;
and merging the sample data table and the pseudo data table to form a new training set, and outputting the training set to the model training unit.
Preferably, the model training unit includes construction of a T-O-W model, where the T-O-W model is an expression relationship model of a temperature value and a dissolved oxygen amount with a generated power, and a construction process of the T-O-W model is as follows:
the temperature value and the dissolved oxygen in each group of data in the training set are used as the output of the RBF neural network model, and each generated power is used as the input to train the RBF neural network model;
finding an approximation functionBy minimizing the following objective function:
，
set training set asIs a training data vector, wherein->Representing the->Data of group->Representing the number of training set data sets, +.>Is->A Gaussian basis function corresponding to the group data;
updating the weight vector, acquiring the mapping relation between input and output, and generating a T-O-W model expressed by the relation between the temperature value and dissolved oxygen and the generated power.
Preferably, the RBF neural network model Loss function is set as a Huber Loss function expressed as,
，
wherein,representing hyper-parameters->Temperature value and dissolved oxygen amount representing data set in training set, +.>The temperature value and the dissolved oxygen amount which are predicted by the RBF neural network model are represented,
and gradually reducing the value of the loss function by calculating the gradient information of the loss function, optimizing the RBF neural network model, enabling the predicted output to continuously approximate to the temperature value and the dissolved oxygen of the real training set, and stopping training when the value of the loss function of the training set is not reduced any more, thus obtaining the trained RBF neural network model.
Preferably, the judging unit is used for analyzing and judging the expected generated power and the real-time generated power, and the generated power at the current moment is measured through the power sampling device and recorded as the real-time generated powerThe power set by the manager is recorded as the desired power>The judging unit carries out numerical judgment through a judging model, and the objective function of the judging model is as follows:
will real-time power generationAnd the desired generation power->Inputting a judgment model according to->And the output value of (2) is subjected to relevant adjustment.
Preferably, the judgment unit pairThe output value judgment result of (a) is as follows:
if it isRepresents real-time generated power outputted from microbial fuel cell MFC +.>Desired generation power desired by manager +.>Equality, output the safe signal;
if it isRepresents real-time generated power outputted from microbial fuel cell MFC +.>Desired generation power desired by manager +.>And if the reaction conditions are not equal, outputting a warning signal, and adjusting the reaction environment in the current cycle time of the microbial fuel cell MFC by an adjusting unit.
Preferably, the adjusting unit adjusts the temperature value and the dissolved oxygen amount in the MFC of the microbial fuel cell based on the PID control model, and the working process of the adjusting unit is as follows:
acquiring desired power generation of microbial fuel cell MFCWill expect the power generation +.>Inputting into the T-O-W model of the model training unit, obtaining the power of the desired generation +.>Under the condition of output, the temperature value and the dissolved oxygen amount required by the microbial fuel cell MFC are recorded as the expected temperature value and the expected dissolved oxygen amount, and the temperature value and the dissolved oxygen amount at the current moment are obtained through a temperature sensor and a dissolved oxygen analyzer and recorded as the real-time temperature value and the real-time dissolved oxygen amount;
and inputting the expected temperature value, the real-time temperature value, the expected dissolved oxygen amount and the real-time dissolved oxygen amount into a PID control model, wherein the PID control model controls the real-time temperature value and the real-time dissolved oxygen amount to approach the expected temperature value and the expected dissolved oxygen amount.
Preferably, the PID control model performs temperature regulation and control through temperature control equipment, and controls the oxygen input flow of the air compressor through a valve and a flowmeter.
Preferably, the dried beancurd stick residual liquid recycling data processing system is applied to a microcomputer and comprises a man-machine interaction interface, page display is carried out on system operation data through the man-machine interaction interface, management personnel carry out functional operation and access on the system through the man-machine interaction interface, communication connection complies with the HTTP protocol, and an integrated information processing system and a multi-client information system structure are adopted.
Compared with the prior art, the invention has the following beneficial effects:
1. according to the invention, the reaction environment of the microbial fuel cell MFC is monitored in real time through the data acquisition unit, so that more accurate battery state information can be obtained, reliable data support is provided for model training of the sample expansion unit and the model training unit, the situation of data analysis errors caused by less sample size is reduced by using the GAN model to expand the sample library, and more basis is provided for data processing in the T-O-W model construction process. The RBF neural network is used for carrying out data training on the expression relation model of the temperature value, the dissolved oxygen amount and the generated power, so that data support and adjustment basis are provided for adjustment of the temperature value and the dissolved oxygen amount when the generated power is adjusted, and effective operation in the regulation and control process is realized.
2. According to the invention, the expected power generation and the real-time power generation are analyzed and judged through the judging unit, whether the microbial fuel cell MFC needs to be subjected to adjustment of the reaction environment or not is obtained, the temperature value and the dissolved oxygen amount in the microbial fuel cell MFC are adjusted through the PID control model of the adjusting unit, so that the real-time temperature value and the real-time dissolved oxygen amount approach to the expected temperature value and the expected dissolved oxygen amount, the effective adjustment of the reaction environment of the microbial fuel cell MFC is realized, the output state of the MFC is effectively controlled, the loss of the cell in a bad state is reduced, and the data are analyzed, so that a better reaction environment is provided for the microbial fuel cell MFC reaction environment, and the performance of the cell can be fully exerted.
Drawings
Fig. 1 is a schematic structural diagram of a dried beancurd stick raffinate recycling data processing system based on data analysis according to the present invention;
FIG. 2 is a schematic diagram of the method of the present invention;
fig. 3 is a schematic diagram of a reaction model of a microbial fuel cell MFC according to the present invention.
Detailed Description
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are only some of the embodiments of the present application and not all of the embodiments of the present application, and it should be understood that the present application is not limited by the example embodiments described herein.
Example 1
Referring to fig. 1, fig. 2 and fig. 3, a data processing system for recovering dried beancurd sticks based on data analysis according to the present invention is further described.
For treating the high-concentration dried beancurd stick wastewater, a plurality of process technologies are generally required to be combined for treatment so as to meet the discharge requirement, but a large amount of chemical agents are required to be added in the operation process, so that the water quality is affected to a certain extent, the operation cost is high due to the addition of the chemical agents, the aerobic aeration and the wastewater reflux, and meanwhile, the management difficulty of the combined process is high. The microbial fuel cell (Microbial Fuel Cell, MFC) is used as a novel wastewater treatment technology, has low requirements on water inflow, strong pollution load resistance and can directly convert chemical energy in wastewater organic matters into electric energy without external energy intake. The high-concentration dried beancurd stick wastewater can provide sufficient carbon sources for the long-term stable operation of the MFC, and is beneficial to the generation of electric energy; and the MFC has good removal effect on organic matters in the wastewater, and has the advantages of cleanness, energy conservation, economy and the like compared with the traditional combined process technology.
The method comprises the steps of constructing a double-chamber type MFC by taking waste liquid of dried beancurd sticks as a substrate, taking the waste liquid of dried beancurd sticks as a microorganism source, calculating the temperature and rapidly analyzing and processing data in the oxygen demand control process in the process of processing load voltage, acquiring the temperature by a temperature sensor, regulating and controlling the temperature by temperature control equipment, monitoring and acquiring the content data of oxygen in the liquid by a dissolved oxygen analyzer, connecting a cathode with an air compressor, and controlling the oxygen input flow by a valve and a flowmeter.
The anode is the main carrier for the growth and propagation of electroactive microorganisms and extracellular electron transfer in MFC, and its physical, chemical and surface properties directly affect MFC performance. The biocompatibility, the high specific surface area and the porosity of the anode are important for supporting the attachment and the fixation of microorganisms, and in order to improve the wastewater treatment effect and the power generation performance of a Microbial Fuel Cell (MFC), the system takes a Carbon Felt (CF) (SA-PQ-11/CF) as the anode, the SA-PQ-11/CF anode has a larger specific surface area, and the solution resistance and the charge transfer resistance of the MFC are obviously reduced. And an MFC system is constructed by taking dried beancurd stick wastewater as an anolyte and CF as a cathode.
The microbial fuel cell MFC is composed of an anode unit and a cathode unit. In the anode unit, the anode is buried in the sediment at the lower layer and is in an anaerobic state, and electroactive bacteria on the anode perform growth metabolism by utilizing organic matters and the like in the sediment and release electrons and protons. The anode acts as an electron acceptor, receives electrons and transfers them to the cathode through an external circuit, and protons diffuse to the cathode unit through a mud/water interface. In the cathode unit, electrons, protons and final electron acceptors undergo a reduction reaction, thereby forming available current, and the electrode reaction formula is as follows:
anode reaction:
cathode reaction:
the sample model building module comprises a data acquisition unit, a sample expansion unit and a model training unit, wherein the data acquisition unit is used for acquiring temperature values and dissolved oxygen amounts of corresponding time sequences in the current period time, acquiring power generation power on the time sequences corresponding to the temperature values and the dissolved oxygen amounts, measuring the power generation output power by a power sampling device to obtain a sample database, forming a sample data table, the sample expansion unit expands the sample data table through a GAN model to generate a pseudo data table, merging the sample data table and the pseudo data table to form a new training set, and the model training unit carries out training according to the training set to acquire an expression relation model of the temperature values, the dissolved oxygen amounts and the power generation.
The sample database comprises a plurality of time sequences marked randomly in each cycle time, a plurality of time points are obtained by the system, and the temperature sensor, the dissolved oxygen analyzer and the electric power sampling device are used for obtaining the temperature of the reaction liquid of the microbial fuel cell MFC, the liquid dissolved oxygen (oxygen dissolved amount) at the cathode and the electric power output by the MFC corresponding to each time point, so that the sample database is formed, each item of data in the sample database is denoised and smoothed and then is recorded into an excel table to form a sample data table, wherein a plurality of temperature sensors and the dissolved oxygen analyzer are arranged on the reaction tank for obtaining the accuracy of the data, and only one temperature value and one dissolved oxygen amount are output at the same time point in the sample database, so that when the data output of the plurality of temperature sensors and the dissolved oxygen analyzer exist, the data can be subjected to average treatment, and only one average temperature value and one average dissolved oxygen amount are output at the same time point.
The molecular oxygen in the air dissolved in water is called dissolved oxygen, and the content of the dissolved oxygen in the water has close relation with the partial pressure of the oxygen in the air and the temperature of the water. In general, the oxygen content in the air does not change much, so the water temperature is a main factor, and the lower the water temperature is, the higher the content of dissolved oxygen in the water is.
Because the sampling times of the temperature sensor, the dissolved oxygen analyzer and the electric sampling device are limited in the period time, or the data quantity obtained in the sample data table is limited in the period time due to the consideration of cost and other factors, the sample data quantity is small, and the relation between the expression relation of the temperature value and the dissolved oxygen quantity and the generating power and the actual deviation are large in the model training process and cannot be well expressed, the model training needs sufficient sample data quantity to achieve the ideal effect.
In the sample expansion unit, the training process of expanding the sample data table by the GAN model and outputting the reference sample is as follows:
constructing a framework of a GAN model: selection generatorDiscriminator->Neural network structure and random noise of (a)Setting the iteration times;
training a GAN model according to the objective function to obtain a trained GAN model, inputting each group of data of the sample data table into the trained GAN model, and setting and generating Q groups of pseudo data tables;
and merging the sample data table and the pseudo data table to form a new training set, and outputting the training set to the model training unit.
The frame is composed of a generator(Generator) and a discriminator->(Generator) the entire training process, namely the two oppositional games: given->，/>Representing the entered real data, < > in the present system>Representing a sample data table, +.>Representing an arbitrary set of data in the sample data table, and the set of data includes the corresponding temperature value, dissolved oxygen amount and power generation at the measurement time point, hopefully learning generator->Discriminator->So that->，/>Representing the generated dummy data set so that the arbiter +.>The two cannot be correctly distinguished.
A generatorCan be any neural network structure, and the input is random noise +.>The random noise can be Gaussian, uniformly distributed, etc., the training goal is to let the output +.>The larger probability can be obtained in the subsequent judging stage; discriminator->Can be any neural network structure, the input is +.>Or->The method comprises the steps of carrying out a first treatment on the surface of the The training goal is to accurately distinguish between the two, giving +.>With greater probability ∈>With a smaller probability (two classes)Two networks train alternately, for the discriminator +.>Multiple rounds of optimization are performed, whereas the generator is +.>Only one round of optimization is performed, and the capacity is synchronously improved.
The training of GAN is a solution to the minmax problem, and the corresponding objective function is as follows:
；
in the method, in the process of the invention,for a true sample distribution, a sample set sampled from a true training sample is represented,representing a real sample randomly sampled from the real sample distribution, i.e., a set of data randomly sampled from a sample data table; />Is a random noise distribution; />Representing a noise sample randomly sampled from a random noise distribution,/for>Representing a neural network function with real samples as variables, < ->Expressed in noise samples->Neural network function as a variable, +.>Expressed as +.>Neural network functions that are variables;
training a GAN model to maximize an objective function; through iterative training, parameters of a generator and a discriminator are continuously adjusted, and a GAN model is finally obtained;
in the invention, each group of data of a sample data table in the current period time is input into a trained GAN model, a global optimal solution is obtained through solving a GAN objective function, Q groups of pseudo data are set and generated, the sample data table and the pseudo data table are combined to form a new training set, and the training set provides data support for a model training unit.
The model training unit comprises construction of a T-O-W model, wherein the T-O-W model is an expression relation model of a temperature value, dissolved oxygen and generated power, and the construction process of the T-O-W model is as follows:
the temperature value and the dissolved oxygen amount in each group of data in the training set are used as the output of the RBF neural network model, each power generation power is used as the input, the RBF neural network model is trained, the weight vector is updated, the mapping relation between the input and the output is obtained, when the power generation power is required to be set, the set power generation power is input into the trained RBF neural network model, and the temperature value and the dissolved oxygen amount corresponding to the power generation power are generated;
in the RBF neural network, the input layer receives input data and transmits the input data to the hidden layer; the hidden layer is composed of a series of radial basis functions that calculate the output from the distance between the input data and the respective center point; each basis function corresponds to a central point and a deviation parameter, and the basis functions can adapt to the characteristics of input data by adjusting the parameters; the output layer calculates a final prediction result according to the output of the hidden layer; typically, the output layer uses a linear activation function to obtain a continuous prediction value.
The invention adopts RBF neural network to predict the relation expression of the generated power, the corresponding temperature value and the dissolved oxygen amount in the period time, solves the generalization problem of the traditional BP neural network, and can obtain better results by adopting a random approximation method:
set training set asIs a training data vector, wherein->Representing the->Data of group->Representing the number of training set data sets, +.>Is the central point of the radial basis function of the network, wherein +.>Indicate->The data of the group corresponds to the center point of the basis function, < >>Representing the number of basis functions corresponding to the data set,/->Is a Gaussian basis function; the basic idea of regularization is to control the smoothness of the mapping function by adding a constraint containing a priori knowledge of the solution, so that similar inputs correspond to similar outputs, finding the approximation function +.>By minimizing the following objective function:
，
the gaussian basis function is:
,
for input to->Group data corresponds to the center vector of the hidden layer, +.>Is->The base width parameter of the node corresponding to the group data, so the network output is +.>,/>Is->Group data corresponding weights, < >>Is->The group data corresponds to a gaussian basis function value; and->,/>Is->Symmetry matrix of the data sets of the training set +.>Is a regularization parameter, +.>Is a correction coefficient.
When the Loss function of the RBF neural network model is set, a Huber Loss function is adopted, and Huber Loss is a Loss function which combines MSE and MAE and takes advantages of the two, and is also called Smooth MeanAbsolute ErrorLoss (Smooth L1 Loss). Huber Loss is also a Loss function used in regression that is not as sensitive to outliers in the data as the square error Loss. It has the characteristics of insensitivity to abnormal points and very small and tiny, so that the loss function has good properties. The MSE portion of Huber Loss is utilized when the error is small, and the MAE portion of Huber Loss is utilized when the error is large. A new super-parameter is introduced that tells the position of the switch of the loss function from MSE to MAE. Introducing super-parameters into the Loss function to enable the transition from MSE to MAE to be smooth, effectively measuring the difference between the predicted output of the model and the true value, huber Loss function,
，
wherein,representing hyper-parameters->Temperature value and dissolved oxygen amount representing data set in training set, +.>The temperature value and the dissolved oxygen amount obtained through RBF neural network model prediction are shown, the Huber Loss enhances the robustness of outliers of MSE, and the sensitivity problem to the outliers is reduced. When the error is large, the use of MAE can reduce outlier effects, making training more robust. The descent speed of the MAE is between the MSE and the MSE, so that the problem of slow descent speed of the MAE in the Loss is solved, and the MAE is closer to the MS.
In the process of training the RBF neural network model by using the training set, gradually reducing the value of the loss function by calculating gradient information of the loss function and updating according to the weight value, so that the predicted output of the RBF neural network model is continuously approximate to the temperature value and dissolved oxygen of the real training set, thereby optimizing the model; gradient clipping technology is adopted to limit the gradient range; terminating training when the value of the loss function of the training set no longer drops; obtaining a trained RBF neural network model.
In this embodiment, the reaction environment of the microbial fuel cell MFC is monitored in real time by the data acquisition unit, so that more accurate battery state information can be obtained, reliable data support is provided for model training of the sample expansion unit and the model training unit, the situation of data analysis errors caused by less sample size is reduced by using the GAN model to expand the sample library, and more basis is provided for data processing in the process of constructing the T-O-W model. The RBF neural network is used for carrying out data training on the expression relation model of the temperature value, the dissolved oxygen amount and the generated power, so that data support and adjustment basis are provided for adjustment of the temperature value and the dissolved oxygen amount when the generated power is adjusted, and effective operation in the regulation and control process is realized.
Example two
Referring to fig. 1 and fig. 2, a second embodiment of the present invention provides a dried beancurd stick raffinate recycling data processing system based on data analysis.
When the power generation power output by the microbial fuel cell MFC needs to be controlled, the judging unit is used for analyzing and judging the expected power generation power and the real-time power generation power, and the power generation power at the current moment is measured through the power sampling device and recorded as the real-time power generation powerThe power generation power required by the manager is obtained and recorded as the desired power generation power +.>The judging unit carries out numerical judgment through a judging model, and the objective function of the judging model is as follows:
will real-time power generationAnd the desired generation power->Inputting a judgment model according to->Is adjusted in dependence on the output value of (a)
If it isRepresents real-time generated power outputted from microbial fuel cell MFC +.>Desired generation power desired by manager +.>The power output at the moment accords with the power generation requirement, additional adjustment is not needed, the current power generation state is kept continuously in the current period time, and a safety signal is output;
if it isRepresents real-time generated power outputted from microbial fuel cell MFC +.>Desired generation power desired by manager +.>The power output at this time is not in accordance with the power generation requirement, the temperature value and dissolved oxygen in the MFC need to be adjusted, a warning signal is output, and the judgment unit is to>And the reaction environment in the current cycle time in the microbial fuel cell MFC is adjusted by the adjusting unit.
The system monitors continuously and periodically, and after each module works in the current period, each module works repeatedly in the next period, namely, at the moment, the judging unit repeatedly generates power in real timeAnd the desired powerIs used for judging the work.
The adjusting unit adjusts the temperature value and the dissolved oxygen amount in the microbial fuel cell MFC based on the PID control model, and the working process of the adjusting unit is that the expected power of the microbial fuel cell MFC is obtainedWill expect power generationInputting into the T-O-W model of the model training unit, obtaining the power of the desired generation +.>Under the condition of output, the temperature value and the dissolved oxygen amount required by the microbial fuel cell MFC are recorded as the expected temperature value and the expected dissolved oxygen amount, the temperature value and the dissolved oxygen amount at the current moment are obtained through a temperature sensor and a dissolved oxygen analyzer in the current period time and recorded as the real-time temperature value and the real-time dissolved oxygen amount, the expected temperature value and the real-time temperature value, the expected dissolved oxygen amount and the real-time dissolved oxygen amount are input into a PID control model, the PID control model carries out temperature regulation and control through temperature control equipment, and the oxygen input flow of the air compressor is controlled through a valve and a flowmeter.
The PID control model approximates and regulates the real-time temperature value and the real-time dissolved oxygen of the microbial fuel cell MFC to the expected temperature value and the expected dissolved oxygen based on the PID control of the neural network, and the construction process of the PID control model is as follows:
the discrete regulatory objective function is constructed and the control system,
formula (1)
Formula (2)
Formula (3)
The formula (1) is used for showing an expression that the gas input speed output by each iteration approaches to the preset input speed;
the formula (2) is used for showing a regulation expression of the iterative gas input speed;
equation (3) is used for showingIs a derived expression of (1);
wherein,representing the current iteration number, +.>Indicating the set expected value, wherein the expected value comprises the expected temperature value and the expected dissolved oxygen amount, +.>Representing the real-time value output in the current iteration number, wherein the real-time value comprises a real-time temperature value and a real-time dissolved oxygen amount, and is simplified to be represented as an iteration output value,/for>The absolute value of the deviation of the iteration output value from the desired value representing the current iteration number,representing the real-time value output in the last iteration number,/-, for example>Representing the difference between the last iteration number and the real value in the current iteration number,/>Representing the absolute value of the deviation of the iteration output value from the expected value in the last iteration number, +.>Absolute value of deviation of iteration output value of last iteration number of the number of iterations t-1 from expected value,/v>Representing an iterative function; />Representing the iteration number; />Is a proportionality coefficient; />Representing integral coefficient>Representing the differential coefficient;
comparing the proportional coefficients based on the real-time values output after each iterationIntegral coefficient->Differential coefficient->And carrying out self-adaptive adjustment, and finally enabling the real-time temperature value and the real-time dissolved oxygen to continuously approach the expected temperature value or the expected dissolved oxygen.
Based on the above process, the PID control model exemplifies the regulation and control of the real-time temperature value and the real-time dissolved oxygen amount, the real-time temperature value and the real-time dissolved oxygen amount are set to be 1 and 2 respectively, and the expected temperature value and the expected dissolved oxygen amount are set to be 3 and 0.5 respectively;
the PID control model regulates the temperature: acquisition of=1，/>=3, i.e. +.>After one time regulation, =2, ++can be obtained>Sequentially repeating until the expected transmission speed is approximately 3, wherein the regulation and control process of the real-time dissolved oxygen is similar to that of the real-time temperature value, and the proportionality coefficient is +.>Integral coefficient->Differential coefficient->Is regulated and controlled by the updating of the neural network model.
In this embodiment, the determination unit analyzes and determines the expected power generation and the real-time power generation to obtain whether the microbial fuel cell MFC needs to adjust the reaction environment, and the PID control model of the adjustment unit adjusts the temperature value and the dissolved oxygen in the microbial fuel cell MFC, so that the real-time temperature value and the real-time dissolved oxygen approach to the expected temperature value and the expected dissolved oxygen, the effective adjustment of the reaction environment of the microbial fuel cell MFC is achieved, the output state of the MFC cell is effectively controlled, the loss of the cell in the bad state is reduced, and by analyzing these data, a better reaction environment is provided for the reaction environment of the microbial fuel cell MFC, and the performance of the cell can be fully exerted.
Example III
Referring to fig. 1 and fig. 2, a second embodiment of the present invention provides a dried beancurd stick raffinate recycling data processing system based on data analysis.
A dried beancurd stick residual liquid recovery data processing system based on data analysis is applied to a microcomputer and comprises a human-computer interaction interface, page display is carried out on system operation data through the human-computer interaction interface, management personnel carry out functional operation and access on the system through the human-computer interaction interface, communication connection complies with the HTTP protocol, and an integrated information processing system and a multi-client information system structure are adopted. The dried beancurd stick raffinate recovery data processing system based on data analysis is applied to a cloud processing platform, is applied to a cloud terminal and performs distributed storage.
The user performs identity verification through a human-computer interface, the filter intercepts login requests of each manager, the authentication gateway checks the digital certificate of the user Ukey with identity information stored in the database, whether the user has login permission is judged, only the manager with the login permission is allowed to log in the system, the manager has permission to access the system after logging in the system, and the human-computer interaction interface can be a computer or other intelligent equipment meeting functions.
The dried beancurd stick raffinate recovery data processing system based on data analysis comprises the following steps:
step S1: constructing a microbial fuel cell MFC reaction model, constructing a double-chamber MFC by taking waste liquid of dried beancurd sticks as a substrate, setting SA-PQ-11/CF as an anode, taking dried beancurd sticks waste water as an anolyte and CF as a cathode, and constructing an MFC power generation system;
step S2: in the current period time, the data acquisition unit acquires and records the power generation power of the fuel cell MFC through the power sampling device, acquires the temperature of the reaction liquid through the temperature sensor, and measures and records the dissolved oxygen amount of the liquid through the dissolved oxygen analyzer to form a sample data table;
step S3: the sample expansion unit expands the sample data table through the GAN model to generate a pseudo data table, and the sample data table and the pseudo data table are combined to form a new training set;
step S4: the model training unit carries out training according to the training set, acquires an expression relation model of the temperature value, the dissolved oxygen and the generated power, and constructs a T-O-W model;
step S5: inputting expected power generation, judging whether the adjusting unit is started or not by the judging unit, if not, ending, and repeating the steps S2 to S5 in the next period;
step S6: if the microbial fuel cell is started, the expected generated power is input into a T-O-W model, an expected temperature value and an expected dissolved oxygen amount are obtained, the adjusting unit adjusts the real-time temperature value and the real-time dissolved oxygen amount in the microbial fuel cell MFC to approach the expected temperature value and the expected dissolved oxygen amount based on the PID control model, and the steps S2 to S6 are repeated in the next period.
Additionally, in accordance with embodiments of the present application, a process described in the figures of a beancurd stick raffinate recovery data processing system based on data analysis may be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium having stored thereon machine-readable instructions executable by a processor to perform the instructions corresponding to the method steps provided herein, of course, the architecture shown in the figures of a beancurd sheet recovery data processing system based on data analysis is merely exemplary, and when different devices are implemented, the adaptation or adjustment may be made according to actual needs.
The above formulas are all formulas with dimensionality removed and numerical calculation, the formulas are formulas with the latest real situation obtained by software simulation through collecting a large amount of data, and preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.
The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above examples, and all technical solutions belonging to the concept of the present invention belong to the protection scope of the present invention. It should be noted that modifications and adaptations to those skilled in the art without departing from the principles of the present invention are intended to be comprehended within the scope of the present invention.
Claims (10)
1. The dried beancurd stick raffinate recovery data processing system based on data analysis is characterized by comprising:
the reaction module is used for constructing an MFC reaction model of the microbial fuel cell, constructing a double-chamber MFC by taking waste liquid of dried beancurd sticks as a substrate, setting SA-PQ-11/CF as an anode, taking dried beancurd sticks waste water as an anolyte and CF as a cathode, and constructing an MFC power generation system;
the sample model building module comprises a data acquisition unit, a sample expansion unit and a model training unit, wherein the data acquisition unit is used for acquiring temperature values and dissolved oxygen amounts of corresponding time sequences in the current period time, acquiring power generation power of each temperature value and corresponding time sequence of the dissolved oxygen amounts, forming a sample data table, the sample expansion unit expands the sample data table through a GAN model, generating a pseudo data table, merging the sample data table and the pseudo data table to form a new training set, and the model training unit performs training according to the training set to acquire an expression relation model of the temperature values, the dissolved oxygen amounts and the power generation power;
the regulation and control module comprises a judging unit and an adjusting unit, wherein the judging unit is used for analyzing and judging expected power generation and real-time power generation and judging whether to start the adjusting unit, and the adjusting unit adjusts the temperature value and the dissolved oxygen in the MFC based on the PID control model.
2. The data analysis-based dried beancurd stick residual liquid recovery data processing system according to claim 1, wherein the power generation output of the microbial fuel cell MFC is measured by a power sampling device, the temperature of reaction liquid is collected and recorded by a temperature sensor, and the dissolved oxygen amount of the liquid is measured and recorded by a dissolved oxygen analyzer.
3. The system for processing dried beancurd sheet waste recovery data based on data analysis of claim 1, wherein the training process of expanding the sample data table and outputting the reference sample by the GAN model in the sample expansion unit is as follows:
constructing a framework of a GAN model: selection generatorDiscriminator->Neural network structure and random noise +.>Setting the iteration times;
training a GAN model according to the objective function to obtain a trained GAN model, inputting each group of data of the sample data table into the trained GAN model, and setting and generating Q groups of pseudo data tables;
and merging the sample data table and the pseudo data table to form a new training set, and outputting the training set to the model training unit.
4. The system for recycling data of dried beancurd sticks based on data analysis according to claim 3, wherein the model training unit comprises construction of a T-O-W model, wherein the T-O-W model is an expression relation model of a temperature value, dissolved oxygen and generated power, and the construction process of the T-O-W model is as follows:
the temperature value and the dissolved oxygen in each group of data in the training set are used as the output of the RBF neural network model, and each generated power is used as the input to train the RBF neural network model;
finding an approximation functionBy minimizing the following objective function:
，
set training set asIs a training data vector, wherein->Representing the->Data of group->Representing the number of training set data sets, +.>Is->A Gaussian basis function corresponding to the group data;
updating the weight vector, acquiring the mapping relation between input and output, and generating a T-O-W model expressed by the relation between the temperature value and dissolved oxygen and the generated power.
5. The system for recovering and processing data from a dried beancurd sheet as recited in claim 4, wherein the RBF neural network model has a Loss function set as a Huber Loss function expressed as,
，
wherein,representing hyper-parameters->Temperature value and dissolved oxygen amount representing data set in training set, +.>The temperature value and the dissolved oxygen amount which are predicted by the RBF neural network model are represented,
and gradually reducing the value of the loss function by calculating the gradient information of the loss function, optimizing the RBF neural network model, enabling the predicted output to continuously approximate to the temperature value and the dissolved oxygen of the real training set, and stopping training when the value of the loss function of the training set is not reduced any more, thus obtaining the trained RBF neural network model.
6. The system for processing the dried beancurd stick residual liquid recovery data based on data analysis according to claim 1, wherein the judging unit is used for analyzing and judging expected power and real-time power, and the power at the current moment is measured through the power sampling device and recorded as real-time powerThe power set by the manager is recorded as the desired power>The judging unit carries out numerical judgment through a judging model, and the objective function of the judging model is as follows:
will real-time power generationAnd the desired generation power->Inputting a judgment model according to->And the output value of (2) is subjected to relevant adjustment.
7. The system for recovering and processing data from dried beancurd sticks based on data analysis of claim 6, wherein said judgment unit is configured to compare said data with said dataThe output value judgment result of (a) is as follows:
if it isRepresents real-time generated power outputted from microbial fuel cell MFC +.>Desired generation power desired by manager +.>Equality, output the safe signal;
if it isRepresents real-time generated power outputted from microbial fuel cell MFC +.>Desired generation power desired by manager +.>And if the reaction conditions are not equal, outputting a warning signal, and adjusting the reaction environment in the current cycle time of the microbial fuel cell MFC by an adjusting unit.
8. The data analysis-based dried beancurd stick residual liquid recovery data processing system according to claim 7, wherein the adjusting unit adjusts the temperature value and the dissolved oxygen amount in the microbial fuel cell MFC based on a PID control model, and the adjusting unit works as follows:
acquiring desired power generation of microbial fuel cell MFCWill expect the power generation +.>Input into the T-O-W model of the model training unit to obtain the expected dataElectric power->Under the condition of output, the temperature value and the dissolved oxygen amount required by the microbial fuel cell MFC are recorded as the expected temperature value and the expected dissolved oxygen amount, and the temperature value and the dissolved oxygen amount at the current moment are obtained through a temperature sensor and a dissolved oxygen analyzer and recorded as the real-time temperature value and the real-time dissolved oxygen amount;
and inputting the expected temperature value, the real-time temperature value, the expected dissolved oxygen amount and the real-time dissolved oxygen amount into a PID control model, wherein the PID control model controls the real-time temperature value and the real-time dissolved oxygen amount to approach the expected temperature value and the expected dissolved oxygen amount.
9. The data analysis-based dried beancurd stick residual liquid recovery data processing system according to claim 8, wherein the PID control model is used for temperature regulation and control through temperature control equipment, and the oxygen input flow of the air compressor is controlled through a valve and a flowmeter.
10. The data analysis-based dried beancurd stick residual liquid recycling data processing system according to claim 1, wherein the dried beancurd stick residual liquid recycling data processing system is applied to a microcomputer and comprises a man-machine interaction interface, page display is carried out on system operation data through the man-machine interaction interface, management personnel carry out functional operation and access on the system through the man-machine interaction interface, communication connection conforms to the HTTP protocol, and an integrated information processing system is adopted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311812985.3A CN117476125B (en) | 2023-12-27 | 2023-12-27 | Dried beancurd stick raffinate recovery data processing system based on data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311812985.3A CN117476125B (en) | 2023-12-27 | 2023-12-27 | Dried beancurd stick raffinate recovery data processing system based on data analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117476125A true CN117476125A (en) | 2024-01-30 |
| CN117476125B CN117476125B (en) | 2024-04-05 |
Family
ID=89639989
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311812985.3A Active CN117476125B (en) | 2023-12-27 | 2023-12-27 | Dried beancurd stick raffinate recovery data processing system based on data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117476125B (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108503041A (en) * | 2018-03-20 | 2018-09-07 | 许昌学院 | A method of handling yuba wastewater using microorganism battery |
| WO2019033636A1 (en) * | 2017-08-16 | 2019-02-21 | 哈尔滨工业大学深圳研究生院 | Method of using minimized-loss learning to classify imbalanced samples |
| CN111832748A (en) * | 2020-08-24 | 2020-10-27 | 西南大学 | Electronic nose width learning method for performing regression prediction on concentration of mixed gas |
| CN112250161A (en) * | 2020-08-29 | 2021-01-22 | 盐城工学院 | Self-adaptive potential controller of microbial electrolysis battery for azo wastewater treatment and control method thereof |
| CN112966740A (en) * | 2021-03-05 | 2021-06-15 | 西安邮电大学 | Small sample hyperspectral image classification method based on core sample adaptive expansion |
| CN114004333A (en) * | 2021-10-26 | 2022-02-01 | 桂林理工大学 | Oversampling method for generating countermeasure network based on multiple false classes |
| CN114036861A (en) * | 2021-10-25 | 2022-02-11 | 中国石油化工股份有限公司 | Three-dimensional temperature field prediction method based on infoGAN |
| CN114881812A (en) * | 2022-06-01 | 2022-08-09 | 南京工程学院 | Building air conditioner load scene generation method, device and equipment based on conditional timing GAN |
| CN115169735A (en) * | 2022-08-02 | 2022-10-11 | 济南大学 | Electroceramic quality prediction method and system based on sample expansion |
| CN115936227A (en) * | 2022-12-15 | 2023-04-07 | 西安交通大学 | Machine learning-based ion permeation power generation prediction system and method |
| CN115983374A (en) * | 2022-12-30 | 2023-04-18 | 广东电网有限责任公司中山供电局 | Cable partial discharge database sample expansion method based on optimized SA-CACGAN |
| CN117076871A (en) * | 2023-10-16 | 2023-11-17 | 南京邮电大学 | Battery fault classification method based on unbalanced semi-supervised countermeasure training framework |
| CN117174971A (en) * | 2023-09-19 | 2023-12-05 | 吉林化工学院 | Self-adaptive control method of microbial fuel cell based on elasticity compensation |
| WO2023236260A1 (en) * | 2022-06-09 | 2023-12-14 | 苏州大学 | Short-term wind speed prediction method based on complex-valued acceleration algorithm |
-
2023
- 2023-12-27 CN CN202311812985.3A patent/CN117476125B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019033636A1 (en) * | 2017-08-16 | 2019-02-21 | 哈尔滨工业大学深圳研究生院 | Method of using minimized-loss learning to classify imbalanced samples |
| CN108503041A (en) * | 2018-03-20 | 2018-09-07 | 许昌学院 | A method of handling yuba wastewater using microorganism battery |
| CN111832748A (en) * | 2020-08-24 | 2020-10-27 | 西南大学 | Electronic nose width learning method for performing regression prediction on concentration of mixed gas |
| CN112250161A (en) * | 2020-08-29 | 2021-01-22 | 盐城工学院 | Self-adaptive potential controller of microbial electrolysis battery for azo wastewater treatment and control method thereof |
| CN112966740A (en) * | 2021-03-05 | 2021-06-15 | 西安邮电大学 | Small sample hyperspectral image classification method based on core sample adaptive expansion |
| CN114036861A (en) * | 2021-10-25 | 2022-02-11 | 中国石油化工股份有限公司 | Three-dimensional temperature field prediction method based on infoGAN |
| CN114004333A (en) * | 2021-10-26 | 2022-02-01 | 桂林理工大学 | Oversampling method for generating countermeasure network based on multiple false classes |
| CN114881812A (en) * | 2022-06-01 | 2022-08-09 | 南京工程学院 | Building air conditioner load scene generation method, device and equipment based on conditional timing GAN |
| WO2023236260A1 (en) * | 2022-06-09 | 2023-12-14 | 苏州大学 | Short-term wind speed prediction method based on complex-valued acceleration algorithm |
| CN115169735A (en) * | 2022-08-02 | 2022-10-11 | 济南大学 | Electroceramic quality prediction method and system based on sample expansion |
| CN115936227A (en) * | 2022-12-15 | 2023-04-07 | 西安交通大学 | Machine learning-based ion permeation power generation prediction system and method |
| CN115983374A (en) * | 2022-12-30 | 2023-04-18 | 广东电网有限责任公司中山供电局 | Cable partial discharge database sample expansion method based on optimized SA-CACGAN |
| CN117174971A (en) * | 2023-09-19 | 2023-12-05 | 吉林化工学院 | Self-adaptive control method of microbial fuel cell based on elasticity compensation |
| CN117076871A (en) * | 2023-10-16 | 2023-11-17 | 南京邮电大学 | Battery fault classification method based on unbalanced semi-supervised countermeasure training framework |
Non-Patent Citations (3)
| Title |
|---|
| 傅宏辉;王友仁;孙灿飞;孙权;: "基于ACGAN的功率变换器参数性故障诊断方法", 机械制造与自动化, no. 06, 20 December 2019 (2019-12-20) * |
| 王敏敏: "基于状态估计的微生物燃料电池控制方法研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅰ辑》, 15 January 2022 (2022-01-15) * |
| 魏泉增等: "微生物燃料电池法处理腐竹加工废水", 《广州化工》, 31 October 2019 (2019-10-31), pages 67 - 69 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117476125B (en) | 2024-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hua et al. | Remaining useful life prediction of PEMFC systems under dynamic operating conditions | |
| CN110555548A (en) | ICS-ELM ultra-short-term wind power prediction method based on data mining original error correction | |
| CN108595892A (en) | Soft-measuring modeling method based on time difference model | |
| CN112782975A (en) | Sewage treatment aeration intelligent control method and system based on deep learning | |
| CN116435557A (en) | Fuel cell thermal management method, device and system based on neural network | |
| CN117476125B (en) | Dried beancurd stick raffinate recovery data processing system based on data analysis | |
| CN111311026A (en) | Runoff nonlinear prediction method considering data characteristics, model and correction | |
| CN110705752A (en) | Sewage BOD real-time prediction method based on ANFIS and mechanism model | |
| CN116865343B (en) | Model-free self-adaptive control method, device and medium for distributed photovoltaic power distribution network | |
| Mesquita et al. | Machine learning applied for metabolic flux‐based control of micro‐aerated fermentations in bioreactors | |
| CN105372995A (en) | Measurement and control method for sewage disposal system | |
| CN111259340B (en) | Saturation load prediction method based on logistic regression | |
| CN111798134B (en) | Method for improving methane yield of sewage treatment plant based on data-driven model | |
| CN115101136A (en) | Large-scale aluminum electrolysis cell global anode effect prediction method | |
| CN114357867A (en) | Primary frequency modulation control method and device based on water turbine simulation intelligent solution | |
| CN113221436A (en) | Soft measurement method for sewage suspended matter concentration based on improved RBF neural network | |
| CN114065621A (en) | Gas compressor characteristic equation coefficient fitting method based on genetic algorithm | |
| CN114149076A (en) | Intelligent debugging system of anaerobic ammonia oxidation sewage treatment system | |
| CN113449789A (en) | Quality control method for monitoring water quality by full-spectrum water quality monitoring equipment based on big data | |
| Jun et al. | Displacement prediction model for concrete dam based on PSO-GBDT | |
| Zhai et al. | Degradation prediction of 65 kW proton exchange membrane fuel cells on city buses using a hybrid approach with the advantage actor-critic method | |
| CN108428915A (en) | A kind of fuel cell exhaust process anode pressure control method based on iterative learning | |
| CN113095534A (en) | Wind power prediction method combining ARIMA and improved Elman neural network | |
| CN114896871B (en) | Photovoltaic output decomposition method based on minimum mutual information | |
| CN115829140B (en) | Wind power generation plant power generation amount prediction method and system based on machine learning |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |