CN111611747B - Online state estimation method and device for hybrid energy storage system - Google Patents
Online state estimation method and device for hybrid energy storage system Download PDFInfo
- Publication number
- CN111611747B CN111611747B CN202010445937.5A CN202010445937A CN111611747B CN 111611747 B CN111611747 B CN 111611747B CN 202010445937 A CN202010445937 A CN 202010445937A CN 111611747 B CN111611747 B CN 111611747B
- Authority
- CN
- China
- Prior art keywords
- layer
- fuzzy
- weight
- output
- super capacitor
- 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.)
- Active
Links
- 238000004146 energy storage Methods 0.000 title claims abstract description 40
- 238000000034 method Methods 0.000 title claims abstract description 25
- 230000008451 emotion Effects 0.000 claims abstract description 79
- 210000004556 brain Anatomy 0.000 claims abstract description 52
- 239000003990 capacitor Substances 0.000 claims abstract description 50
- 238000000605 extraction Methods 0.000 claims abstract description 21
- 238000005070 sampling Methods 0.000 claims abstract description 10
- 230000006870 function Effects 0.000 claims description 47
- 230000001953 sensory effect Effects 0.000 claims description 43
- 238000013528 artificial neural network Methods 0.000 claims description 35
- 238000012549 training Methods 0.000 claims description 26
- 238000012360 testing method Methods 0.000 claims description 23
- 210000001061 forehead Anatomy 0.000 claims description 22
- 239000013598 vector Substances 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 17
- 235000011437 Amygdalus communis Nutrition 0.000 claims description 16
- 241000220304 Prunus dulcis Species 0.000 claims description 16
- 235000020224 almond Nutrition 0.000 claims description 16
- 210000001103 thalamus Anatomy 0.000 claims description 14
- 238000000354 decomposition reaction Methods 0.000 claims description 13
- 210000004727 amygdala Anatomy 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 9
- 230000005284 excitation Effects 0.000 claims description 8
- 230000001054 cortical effect Effects 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 7
- 238000013139 quantization Methods 0.000 claims description 7
- 238000000513 principal component analysis Methods 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 4
- 230000036541 health Effects 0.000 claims description 2
- 238000003062 neural network model Methods 0.000 abstract 1
- 230000002996 emotional effect Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002457 bidirectional effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 210000005153 frontal cortex Anatomy 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 125000004432 carbon atom Chemical group C* 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000000840 electrochemical analysis Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- 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
- Y02E70/00—Other energy conversion or management systems reducing GHG emissions
- Y02E70/30—Systems combining energy storage with energy generation of non-fossil origin
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Software Systems (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Entrepreneurship & Innovation (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Public Health (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Feedback Control In General (AREA)
Abstract
The invention belongs to the technical field of clean energy storage, and discloses an online state estimation method for a hybrid energy storage system, which comprises the steps of obtaining the instant voltage and the instant current of a storage battery and a super capacitor which are processed by a sampling circuit; when the increase value of the instant current is detected to exceed a preset threshold value, performing online capacity estimation on the instant voltages of the storage battery and the super capacitor; performing feature extraction on the instant voltage, and estimating the residual electric quantity of the storage battery and the super capacitor through a trained fuzzy brain emotion learning model; sending output data of the fuzzy brain emotion learning model; the device is also disclosed, the voltage and the current of the hybrid energy storage system after the load is impacted are utilized, and a neural network model designed by combining a fuzzy inference system and a brain emotion learning model is used as a state classifier, so that the state of the hybrid energy storage system is quickly, accurately and timely estimated on line, the device is suitable for being applied to a micro-grid and an uninterruptible power system, and the reliability of a power system is improved.
Description
Technical Field
The invention belongs to the technical field of clean energy storage, and particularly relates to an online state estimation method and device for a hybrid energy storage system.
Background
With the wide application of clean energy power generation technologies represented by solar energy, wind energy and the like, in order to overcome the volatility of distributed power generation and the influence of external impact load on the safety and reliability of a microgrid, an important support technology made by an energy storage technology is widely valued and involved. The hybrid energy storage system can effectively exert the characteristics of the storage battery and the super capacitor, so that the power supply system can better supply power to the impact load, the internal loss of the storage battery is reduced, and the service life of the storage battery is prolonged. The super capacitor belongs to an emerging energy storage element, and the current research mainly focuses on the analysis of internal parameters and design and manufacture.
However, the conventional energy storage detection method is mostly performed in a regular off-line detection mode, so that a lot of manpower and time are consumed, and the risk of power grid operation is possibly increased; meanwhile, in recent years, parameter modeling analysis of the energy storage module is performed, and online monitoring of the energy storage module is performed according to signals such as voltage and current, but practical application of the energy storage module is difficult due to a complex electrochemical effect inside the energy storage element.
Disclosure of Invention
The invention aims to provide an online state estimation method and device for a hybrid energy storage system, which are used for solving the problems that the conventional energy storage detection method is combined and is mostly carried out in a regular offline detection mode, a large amount of manpower and time are consumed, and the risk of power grid operation is possibly increased; meanwhile, in recent years, parameter modeling analysis of the energy storage module is performed, and online monitoring of the energy storage module is performed according to signals such as voltage and current, but practical application of the energy storage module is difficult due to a complex electrochemical effect inside the energy storage element.
The invention provides an online state estimation method for a hybrid energy storage system, which is used for solving the technical problem and comprises the following steps:
acquiring the instant voltage and the instant current processed by the sampling circuit for the storage battery and the super capacitor;
when the increase value of the instant current is detected to exceed a preset threshold value, performing online capacity estimation on the instant voltages of the storage battery and the super capacitor;
performing feature extraction on the instant voltage, and estimating the residual electric quantity of the storage battery and the super capacitor through a trained fuzzy brain emotion learning model;
and sending output data of the fuzzy brain emotion learning model.
Further preferably, the "performing feature extraction on the instantaneous voltage" specifically includes: extracting effective data characteristics by using a preferred wavelet packet decomposition method, wherein the wavelet basis function family selects Daubechies wavelet family; calculating the band energy entropy of the jth-order Daubechies basis function after the decomposition of the fourth layer of wavelet packets, selecting the order with the minimum band energy entropy as the optimal wavelet basis function, calculating the signal energy corresponding to the band of the fourth layer under the optimal wavelet basis function, and obtaining an energy feature vector through normalization; and taking the energy coefficient of the low-frequency part obtained by decomposing the wavelet packet and the high-frequency energy coefficient which accounts for more than 95% of the total contribution rate after principal component analysis as input characteristic data.
Further preferably, the 'fuzzy brain emotion learning model' is a fuzzy brain emotion learning neural network comprising a characteristic input layer, a sensory cortex layer, a sensory weight layer, an emotion weight layer, a thalamus, an amygdala, a forehead cortex layer and an output layer.
Further preferably, the fuzzy brain emotion learning neural network specifically comprises: the first layer is a fault characteristic input layer, and voltage state characteristics of a storage battery and a super capacitor are introduced into a neural network; the second layer is a sensory cortex layer, and the input characteristic vectors are subjected to quantization processing by taking a Gaussian function as an excitation function; the third layer is a feeling weight layer, an emotion weight layer and a thalamus, namely a weight space, and the fuzzy output of each block represents a partial result of a fuzzy inference rule; the fourth layer is an almond kernel and a forehead cortex layer, and the sensory cortex with the sensory weight and the emotion weight is output to calculate an algebraic sum and is normalized; and the fifth layer is an output layer, the output of the almond kernel and the output of the forehead cortical layer are subtracted, then a state label is output, and the states of the storage battery and the super capacitor are judged through the label.
Further preferably, "performing feature extraction on the instantaneous voltage, and estimating the remaining electric quantity of the storage battery and the super capacitor by the trained fuzzy brain emotion learning model" specifically includes: dividing the characteristic data into a training sample and a test sample; training the fuzzy brain emotion learning neural network state estimator by adopting a training sample and adjusting parameters; when the training error is detected to be in accordance with the expected error value, entering the next step, otherwise, continuously updating the parameters of the feeling cortex layer, the feeling weight layer and the emotion weight layer, and returning to the previous step; and judging whether the correctness of the test sample and the obtained weight test set meets the requirement, and if so, ending the test.
Another technical solution adopted to solve the technical problem of the present invention is to provide an online state estimation device for a hybrid energy storage system, including the following:
the sampling module is used for acquiring the instant voltage and the instant current processed by the sampling circuit for the storage battery and the super capacitor;
the detection module is used for performing online capacity estimation on the instant voltage of the storage battery and the super capacitor when the increase value of the instant current exceeds a preset threshold value;
the data processing module is used for performing feature extraction on the instant voltage and estimating the residual electric quantity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model;
and the communication module is used for sending the output data of the fuzzy brain emotion learning model to maintenance personnel.
Further preferably, the "performing feature extraction on the instantaneous voltage" in the data processing module specifically includes: extracting effective data characteristics by using a preferred wavelet packet decomposition method, wherein the wavelet basis function family selects Daubechies wavelet family; calculating the band energy entropy of the jth-order Daubechies basis function after the decomposition of the fourth layer wavelet packet, selecting the order with the minimum band energy entropy as the optimal wavelet basis function, calculating the signal energy corresponding to the fourth layer frequency band under the optimal wavelet basis function, and obtaining an energy feature vector through normalization; and taking the energy coefficient of the low-frequency part obtained by decomposing the wavelet packet and the high-frequency energy coefficient which accounts for more than 95% of the total contribution rate after principal component analysis as input characteristic data.
Further preferably, the 'fuzzy brain emotion learning model' in the data processing module is a fuzzy brain emotion learning neural network comprising a characteristic input layer, a sensory cortex layer, a sensory weight layer, an emotion weight layer, a thalamus, an amygdala, a forehead cortex layer and an output layer.
Further preferably, the fuzzy brain emotion learning neural network specifically includes: the first layer is a fault characteristic input layer, and voltage state characteristics of a storage battery and a super capacitor are introduced into a neural network; the second layer is a sensory cortex layer, and the input characteristic vectors are subjected to quantization processing by taking a Gaussian function as an excitation function; the third layer is a feeling weight layer, an emotion weight layer and a thalamus, namely a weight space, and the fuzzy output of each block represents a partial result of a fuzzy inference rule; the fourth layer is an almond kernel and a forehead cortex, and the sensory cortex with the sensory weight and the emotion weight is output to calculate algebraic sum and is normalized; and the fifth layer is an output layer, the output of the almond kernel and the output of the forehead cortical layer are subtracted, then a state label is output, and the states of the storage battery and the super capacitor are judged through the label.
Further preferably, the "performing feature extraction on the instantaneous voltage, and estimating the remaining capacity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model" specifically includes: dividing the characteristic data into a training sample and a test sample; training the fuzzy brain emotion learning neural network state estimator by adopting a training sample and adjusting parameters; when the training error is detected to be in accordance with the expected error value, entering the next step, otherwise, continuously updating the parameters of the feeling cortex layer, the feeling weight layer and the emotion weight layer, and returning to the previous step; and judging whether the correctness of the test sample and the obtained weight test set meets the requirement, and if so, ending the test.
The invention has the beneficial effects that:
1. the invention does not need to carry out complicated electrochemical analysis on the storage battery and the super capacitor and also does not need to detect the high-frequency harmonic signals injected into the storage battery and the super capacitor; under the action of an impact load, the voltage and the current of the storage battery and the super capacitor are corresponding, so that the health state of the hybrid energy storage system can be quickly, timely and accurately estimated, the hybrid energy storage system is suitable for an uninterruptible power supply system and an energy storage system of a transformer substation, the system maintenance cost is reduced, and the reliability of the system is improved;
2. the fuzzy brain emotion learning neural network combines the fuzzy system theory and the neural network with emotion intelligence, so that the fuzzy brain emotion learning neural network has the characteristics of high convergence rate, low calculation complexity and the like, and can meet the requirements of engineering.
Drawings
Fig. 1 is a block flow diagram of an online state estimation method for a hybrid energy storage system according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a hybrid energy storage system structure and an online state estimator in an online state estimation method for a hybrid energy storage system according to an embodiment of the present invention;
fig. 3 is a flowchart of a feature extraction algorithm in an online state estimation method for a hybrid energy storage system according to an embodiment of the present invention;
fig. 4 is a flowchart of offline training and online estimation of FBELNN in an online state estimation method for a hybrid energy storage system according to an embodiment of the present invention;
fig. 5 is a schematic network structure diagram of FBELNN in the online state estimation method for the hybrid energy storage system according to the embodiment of the present invention;
fig. 6 is a schematic structural diagram of an online state estimation device for a hybrid energy storage system according to an embodiment of the present invention.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention and/or the technical solutions in the prior art, the following description will explain specific embodiments of the present invention with reference to the accompanying drawings. It is obvious that the drawings in the following description are only some examples of the invention, and that for a person skilled in the art, other drawings and embodiments can be derived from them without inventive effort. In addition, the reference to the orientation merely indicates the relative positional relationship between the respective members, not the absolute positional relationship.
Referring to fig. 1, fig. 2, fig. 3, fig. 4 and fig. 5, the online state estimation method for a hybrid energy storage system of the present embodiment includes the following steps:
and S1, acquiring the instant voltage and the instant current processed by the sampling circuit for the storage battery and the super capacitor.
As shown in FIG. 2, the super capacitor and the storage battery can be obtained from commercial power and other renewable energy sources through a bidirectional direct current circuit and a bidirectional rectification circuit, the capacity is divided into 100% -70% which are averagely divided into 15 types, the equivalent impedance is 100% -170%, and the impact power is 150KW-200 KW.
S2, when the increase value of the instantaneous current is detected to exceed the preset threshold value, the online capacity estimation is carried out on the instantaneous voltage of the storage battery and the super capacitor.
And judging whether the impact load occurs or not according to whether the instantaneous current increase value exceeds a preset threshold value or not, and if so, carrying out online real-time estimation according to voltage data of the storage battery and the super capacitor 1 second before and after the impact occurs.
S3, performing feature extraction on the instantaneous voltage, and estimating the residual electric quantity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model.
As shown in fig. 3, the "performing feature extraction on the instantaneous voltage" may specifically include: extracting effective data characteristics by using a preferred wavelet packet decomposition method, wherein the wavelet packet decomposition is divided into four layers, and a Daubechies wavelet family is selected as the wavelet basis function family; calculating the band energy entropy of the jth-order Daubechies basis function after the decomposition of the fourth layer wavelet packet, selecting the basis function with the minimum band energy entropy as the optimal wavelet basis function according to the information entropy theory, calculating the signal energy corresponding to the fourth layer frequency band under the optimal wavelet basis function, and obtaining an energy feature vector through normalization; the calculation formula of the nth-order frequency band energy entropy is as follows:
in the above formula, the first and second carbon atoms are,
the energy probability density of the frequency band i after the fourth layer of wavelet packets are decomposed;
then, selecting the minimum value of the frequency band energy entropy, and calculating according to the following formula:
and m is the order of the optimal vanishing moment, so that the basis function of the optimal wavelet packet can be determined.
Under the basis function of the optimal wavelet packet, the energy of the ith frequency band in the jth layer is:
wherein the content of the first and second substances,
the coefficients are reconstructed for each node of the wavelet packet.
The total signal energy can then be calculated as follows:
the information energy and normalization can be calculated as follows:
the energy feature vector may be calculated as follows:
and taking the energy characteristic vector of the low-frequency part obtained by decomposing the wavelet packet and the high-frequency energy characteristic vector which accounts for more than 95% of the total contribution rate after principal component analysis as input characteristic data.
It should be noted that the 'fuzzy brain emotion learning model' is a fuzzy brain emotion learning neural network comprising a characteristic input layer, a sensory cortex layer, a sensory weight layer, an emotion weight layer, a thalamus, an amygdala, a forehead cortex layer and an output layer; the fuzzy brain emotion learning neural network can be referred to fig. 5, and specifically includes: the first layer is a fault characteristic input layer, and voltage state characteristics of a storage battery and a super capacitor are introduced into a neural network; the second layer is a sensory cortex layer, and the input characteristic vectors are subjected to quantization processing by taking a Gaussian function as an excitation function; the third layer is a feeling weight layer, an emotion weight layer and a thalamus, namely a weight space, and the fuzzy output of each block represents a partial result of a fuzzy inference rule; the fourth layer is an almond kernel and a forehead cortex, and the sensory cortex with the sensory weight and the emotion weight is output to calculate algebraic sum and is normalized; and the fifth layer is an output layer, the output of the almond kernel and the output of the forehead cortical layer are subtracted, then a state label is output, and the states of the storage battery and the super capacitor are judged through the label.
Wherein the first layer is a fault characteristic input layer: and introducing the voltage state characteristics of the storage battery and the super capacitor into the neural network. The method comprises the following specific steps: sample feature vectors
Sending the input into a fuzzy brain emotion learning neural network for forward calculation;
the second layer is the sensory cortex layer: taking a Gaussian function as an excitation function, carrying out quantization processing on the input characteristic vector and transmitting the input characteristic vector to a weight space; each block performs fuzzy excitation on the input features due to the need to improve the generalization capability and the running speed of the model. The Gaussian function is adopted as the activation function and can be expressed as
Expressed in the above equation as the output represented on the jth block for the ith input feature,
and
respectively expressed as Gaussian functionsMean and variance of (c).
The third layer is the sensory weight layer, emotional weight layer and thalamus: that is, the weight space, the fuzzy output of each block represents partial result of the fuzzy inference rule, the sensory cortex output value and the sensory weight multiplication value are sent to the almond kernel,
is composed of
The corresponding perceptual weights are given by:
the sensory cortex output value and emotional weight vector multiplication value are fed into the forehead cortex,
is composed of
The corresponding perceptual weights are given by:
the thalamus plays an inhibitory role on amygdala, and the Vth parameter is set artificially:
the fourth layer is the almond and the forehead cortex: and outputting the sensory cortex with the sensory weight and the emotion weight to calculate algebraic sum, and normalizing. The corresponding output of the almond kernel is as follows:
wherein the forehead cortex corresponding output is:
the fifth layer is an output layer: subtracting the output of the almond kernel and the forehead cortical layer, compressing the output between (0, 1) through a sigmoid function, and judging the states of the storage battery and the super capacitor through the label. Wherein the output Ui is:
in the present embodiment, for the perceptual weight
And emotional weight
Based on neurophysiological principles, the amygdala responds to specific emotional cues and the output of the frontal cortex is adjusted to minimize the difference between the amygdala and the emotional cues. Therefore, can obtain
The update method of (1) is as follows:
wherein, the first and the second end of the pipe are connected with each other,
is that
The learning rate of (a) is determined,and reward the signal
Is that
Consists of an output error signal and an output signal of a neural network.
The expression of (a) is as follows:
wherein, the first and the second end of the pipe are connected with each other,
and
the adjustment is carried out by a human being,
in order to output the error, the error is output,
is the desired output;
in order to inhibit or modulate the signal of the amygdala to accelerate the learning process in the frontal cortex,
the update method of (1) is as follows:
in general, the sensory cortex of the traditional brain emotion learning neural network has no studyThe process is learned, but in FBELNN the updating of the mean and variance of the gaussian function needs to be considered, which makes the sensory cortex have learning rules. Here, the adjustment is made by means of a gradient descent method
And
parameters, obtained by the chain derivative rule:
wherein the content of the first and second substances,
in particular, "S3 performs feature extraction on the instantaneous voltage, and estimates the remaining power of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model" as shown in fig. 4, which specifically includes: s31 dividing the characteristic data into training samples and testing samples; s32, training the fuzzy brain emotion learning neural network state estimator by adopting the training sample and adjusting parameters; s33, when detecting that the training error is in accordance with the expected error value, entering step S34, otherwise, continuing to update the parameters of the sensory cortex layer, the sensory weight layer and the emotion weight layer, and returning to step S32; s34 judges whether the accuracy of the test sample and the weight test set obtained in step S33 meets the requirement, if yes, the current step is ended, otherwise, the step S32 is returned.
The optimal fuzzy brain emotion learning neural network after offline training and testing can be written into a DSP with a feature processing algorithm and a fuzzy brain emotion learning model, namely, the states of the storage battery and the super capacitor are analyzed in real time.
S4, sending output data of the fuzzy brain emotion learning model; the output result of the fuzzy brain emotion learning model can be transmitted to a maintainer through the communication module, so that the maintainer can conveniently judge the situation and perform subsequent maintenance.
According to the embodiment, the impact load response signals of the storage battery and the super capacitor are selected as the diagnosis basis, the used feature extraction method is more intelligent, and the steps of wavelet functions selected through artificial experience and the number of obtained energy features are omitted. The neural network with the emotion module has the characteristics of high convergence rate, low calculation complexity and the like, and can meet the requirements of engineering.
Referring to fig. 6, the present invention further provides an online state estimation device for a hybrid energy storage system, including:
the sampling module is used for acquiring the instant voltage and the instant current processed by the sampling circuit for the storage battery and the super capacitor;
the detection module is used for performing online estimation on the instant voltages of the storage battery and the super capacitor when the instant current is detected to exceed a preset threshold;
the data processing module is used for performing feature extraction on the instant voltage and estimating the residual electric quantity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model;
and the communication module is used for sending the output data of the fuzzy brain emotion learning model to maintenance personnel.
In particular, the "performing feature extraction on the instantaneous voltage" in the data processing module specifically includes: extracting effective data characteristics by using a preferred wavelet packet decomposition method, wherein the wavelet basis function family selects Daubechies wavelet family; calculating the band energy entropy of the jth-order Daubechies basis function after the decomposition of the fourth layer wavelet packet, selecting the basis function with the minimum band energy entropy as the optimal wavelet basis function, calculating the signal energy corresponding to the fourth layer frequency band under the optimal wavelet basis function, and obtaining an energy feature vector through normalization; and taking the energy coefficient of the low-frequency part obtained by decomposing the wavelet packet and the high-frequency energy coefficient which accounts for more than 95% of the total contribution rate after principal component analysis as input characteristic data.
In particular, the fuzzy brain emotion learning model in the data processing module is a fuzzy brain emotion learning neural network comprising a characteristic input layer, a sensory cortex layer, a sensory weight layer, an emotion weight layer, a thalamus, an amygdala, a forehead cortex layer and an output layer.
In particular, the fuzzy brain emotion learning neural network specifically comprises: the first layer is a fault characteristic input layer, and voltage state characteristics of a storage battery and a super capacitor are introduced into a neural network; the second layer is a sensory cortex layer, and the input characteristic vectors are subjected to quantization processing by taking a Gaussian function as an excitation function; the third layer is a feeling weight layer, an emotion weight layer and a thalamus, namely a weight space, and the fuzzy output of each block represents a partial result of a fuzzy inference rule; the fourth layer is an almond kernel and a forehead cortex layer, and the sensory cortex with the sensory weight and the emotion weight is output to calculate an algebraic sum and is normalized; and the fifth layer is an output layer, the output of the almond kernel and the output of the forehead cortical layer are subtracted, then a state label is output, and the states of the storage battery and the super capacitor are judged through the label.
Specifically, the "performing feature extraction on the instantaneous voltage, and estimating the remaining power of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model" specifically includes: dividing the characteristic data into a training sample and a test sample; training the fuzzy brain emotion learning neural network state estimator by adopting a training sample and adjusting parameters; when the training error is detected to be in accordance with the expected error value, entering the next step, otherwise, continuously updating the parameters of the feeling cortex layer, the feeling weight layer and the emotion weight layer, and returning to the previous step; and judging whether the correctness of the test sample and the obtained weight test set meets the requirement, and if so, ending the test.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.
Claims (4)
1. An online state estimation method for a hybrid energy storage system, characterized by comprising the steps of:
s1, acquiring the real-time voltage and the real-time current of the storage battery and the super capacitor after being processed by the sampling circuit;
s2, when detecting that the increase value of the instant current exceeds the preset threshold value, performing online capacity estimation through the instant voltages of the storage battery and the super capacitor;
s3, performing feature extraction on the instantaneous voltage, and estimating the residual capacity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model;
s4, sending output data of the fuzzy brain emotion learning model;
wherein, the 'fuzzy brain emotion learning model' in the step S3 is a fuzzy brain emotion learning neural network comprising a characteristic input layer, a sensory cortex layer, a sensory weight layer, an emotion weight layer, a thalamus, an amygdala, a forehead cortex layer and an output layer;
the fuzzy brain emotion learning neural network specifically comprises the following steps: the first layer is a fault characteristic input layer, and voltage state characteristics of a storage battery and a super capacitor are introduced into a neural network; the second layer is a sensory cortex layer, and the input characteristic vectors are subjected to quantization processing by taking a Gaussian function as an excitation function; the third layer is a feeling weight layer, an emotion weight layer and a thalamus, namely a weight space, and the fuzzy output of each block represents a partial result of a fuzzy inference rule; the fourth layer is an almond kernel and a forehead cortex layer, the sensory cortex layer with the sensory weight and the emotion weight is output to calculate an algebraic sum, and normalization is carried out; the fifth layer is an output layer, the output of the almond kernel and the output of the forehead cortical layer are subtracted, then a state label is output, and the states of the storage battery and the super capacitor are judged through the label;
wherein, the step of performing feature extraction on the instantaneous voltage and estimating the residual electric quantity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model in the step S3 specifically comprises the steps of: s31 dividing the characteristic data into training samples and testing samples; s32, training the fuzzy brain emotion learning neural network state estimator by adopting the training sample and adjusting parameters; s33, when detecting that the training error is in accordance with the expected error value, entering step S34, otherwise, continuing to update the parameters of the sensory cortex layer, the sensory weight layer and the emotion weight layer, and returning to step S32; s34 judges whether the accuracy of the test sample and the weight test set obtained in step S33 meets the requirement, if yes, the current step is ended, otherwise, the step S32 is returned.
2. The online state estimation method for the hybrid energy storage system according to claim 1, wherein the step S3 of "performing feature extraction on the instantaneous voltage" specifically includes: extracting effective data characteristics by utilizing a wavelet packet decomposition method, wherein a Daubechies wavelet family is selected as a wavelet basis function family; calculating the band energy entropy of the jth-order Daubechies basis function after the decomposition of the fourth layer wavelet packet, selecting the order with the minimum band energy entropy as the optimal wavelet basis function, calculating the signal energy corresponding to the fourth layer frequency band under the optimal wavelet basis function, and obtaining an energy feature vector through normalization; and taking the energy coefficient of the low-frequency part obtained by decomposing the wavelet packet and the high-frequency energy coefficient which accounts for more than 95% of the total contribution rate after principal component analysis as input characteristic data.
3. An online state estimation device for a hybrid energy storage system, comprising
The sampling module is used for acquiring the instant voltage and the instant current processed by the sampling circuit for the storage battery and the super capacitor;
the detection module is used for performing online estimation on the instant voltages of the storage battery and the super capacitor when the increase value of the instant current is detected to exceed a preset threshold value;
the data processing module is used for performing feature extraction on the instant voltage and estimating the residual capacity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning neural network;
the communication module is used for sending the output data of the fuzzy brain emotion learning neural network to a maintainer;
the fuzzy brain emotion learning model in the data processing module is a fuzzy brain emotion learning neural network comprising a characteristic input layer, a sensory cortex layer, a sensory weight layer, an emotion weight layer, a thalamus, an amygdala, a forehead cortex layer and an output layer;
the fuzzy brain emotion learning neural network specifically comprises the following steps: the first layer is a fault characteristic input layer, and voltage state characteristics of a storage battery and a super capacitor are introduced into a neural network; the second layer is a sensory cortex layer, and the input characteristic vectors are subjected to quantization processing by taking a Gaussian function as an excitation function; the third layer is a feeling weight layer, an emotion weight layer and a thalamus, namely a weight space, and the fuzzy output of each block represents a partial result of a fuzzy inference rule; the fourth layer is an almond kernel and a forehead cortex layer, the sensory cortex layer with the sensory weight and the emotion weight is output to calculate an algebraic sum, and normalization is carried out; the fifth layer is an output layer, the output of the almond kernel and the output of the forehead cortical layer are subtracted, then a state label is output, and the health states of the storage battery and the super capacitor are judged through the label;
the step of performing feature extraction on the instantaneous voltage and estimating the residual electric quantity of the storage battery and the super capacitor through the trained fuzzy brain emotion learning model specifically comprises the following steps of: dividing the characteristic data into a training sample and a test sample; training the fuzzy brain emotion learning neural network state estimator by adopting a training sample and adjusting parameters; when the training error is detected to be in accordance with the expected error value, entering the next step, otherwise, continuously updating the parameters of the feeling cortex layer, the feeling weight layer and the emotion weight layer, and returning to the previous step; and judging whether the correctness of the test sample and the obtained weight test set meets the requirement, and if so, ending the test.
4. The online state estimation device for the hybrid energy storage system according to claim 3, wherein the step of performing feature extraction on the instantaneous voltage in the data processing module specifically comprises: extracting effective data characteristics by utilizing a wavelet packet decomposition method, wherein a Daubechies wavelet family is selected as a wavelet basis function family; calculating the band energy entropy of the jth-order Daubechies basis function after the decomposition of the fourth layer of wavelet packets, selecting the order with the minimum band energy entropy as the optimal wavelet basis function, calculating the signal energy corresponding to the band of the fourth layer under the optimal wavelet basis function, and obtaining an energy feature vector through normalization; and taking the energy coefficient of the low-frequency part obtained by decomposing the wavelet packet and the high-frequency energy coefficient which accounts for more than 95% of the total contribution rate after principal component analysis as input characteristic data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010445937.5A CN111611747B (en) | 2020-05-25 | 2020-05-25 | Online state estimation method and device for hybrid energy storage system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010445937.5A CN111611747B (en) | 2020-05-25 | 2020-05-25 | Online state estimation method and device for hybrid energy storage system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111611747A CN111611747A (en) | 2020-09-01 |
| CN111611747B true CN111611747B (en) | 2022-09-09 |
Family
ID=72203847
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010445937.5A Active CN111611747B (en) | 2020-05-25 | 2020-05-25 | Online state estimation method and device for hybrid energy storage system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111611747B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114114052A (en) * | 2021-11-25 | 2022-03-01 | 福州大学 | Method for rapidly estimating SOH and SOC of battery based on shock response characteristic |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108958037A (en) * | 2018-08-15 | 2018-12-07 | 厦门理工学院 | WAVELET FUZZY brain emotion learning control method, device, equipment and storage medium |
| CN109870654A (en) * | 2019-02-02 | 2019-06-11 | 福州大学 | The online method for dynamic estimation of accumulator capacity based on impact load response characteristic |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9568931B2 (en) * | 2013-06-19 | 2017-02-14 | Nec Corporation | Multi-layer control framework for an energy storage system |
-
2020
- 2020-05-25 CN CN202010445937.5A patent/CN111611747B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108958037A (en) * | 2018-08-15 | 2018-12-07 | 厦门理工学院 | WAVELET FUZZY brain emotion learning control method, device, equipment and storage medium |
| CN109870654A (en) * | 2019-02-02 | 2019-06-11 | 福州大学 | The online method for dynamic estimation of accumulator capacity based on impact load response characteristic |
Non-Patent Citations (3)
| Title |
|---|
| Online Health Estimate of Hybrid Energy Storage System Based on Fuzzy Brain Emotional Learning Neural Networks;Zhifan Xu 等;《2020 International Conference on System Science and Engineering (ICSSE)》;20201117;全文 * |
| 基于模糊神经网络的移动机器人轨迹跟踪控制的研究;吴秋霞;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20181015;全文 * |
| 基于粒子群的模糊神经网络铅酸蓄电池SOC估计;李涛等;《电源技术》;20170120(第01期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111611747A (en) | 2020-09-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Data-based line trip fault prediction in power systems using LSTM networks and SVM | |
| CN109635928B (en) | Voltage sag reason identification method based on deep learning model fusion | |
| CN106251059B (en) | Cable state evaluation method based on probabilistic neural network algorithm | |
| CN108802525A (en) | Equipment fault intelligent Forecasting based on small sample | |
| CN112069727B (en) | Intelligent transient stability evaluation system and method with high reliability for power system | |
| CN111880044B (en) | Online fault positioning method for distribution network containing distributed power supply | |
| CN107807860B (en) | Power failure analysis method and system based on matrix decomposition | |
| CN108535572B (en) | Metering system secondary circuit monitoring method and device based on fundamental wave zero sequence characteristics | |
| CN111160241B (en) | Power distribution network fault classification method, system and medium based on deep learning | |
| CN109870654B (en) | Storage battery capacity online dynamic estimation method based on impact load response characteristic | |
| CN104318485A (en) | Power transmission line fault identification method based on nerve network and fuzzy logic | |
| CN111611747B (en) | Online state estimation method and device for hybrid energy storage system | |
| Yan et al. | A simplified current feature extraction and deployment method for DC series arc fault detection | |
| CN114581699A (en) | Transformer state evaluation method based on deep learning model in consideration of multi-source information | |
| CN116070140B (en) | Power distribution substation safe operation state monitoring system and method | |
| CN116523681A (en) | Load decomposition method and device for electric automobile, electronic equipment and storage medium | |
| Wu et al. | Identification and correction of abnormal measurement data in power system based on graph convolutional network and gated recurrent unit | |
| CN112308208B (en) | Transformer fault diagnosis method based on deep learning model | |
| CN110441725A (en) | Intelligent electric meter method for diagnosing faults and device based on Adaptive Neural-fuzzy Inference | |
| Chu et al. | A relaxed support vector data description algorithm based fault detection in distribution systems | |
| Tan et al. | Application of self-supervised learning in non-intrusive load monitoring | |
| Wang | Remaining Useful Life Prediction of Proton Exchange Membrane Fuel Cell Based on Deep Learning | |
| Jharbade et al. | Fault Location Estimation for a Bipolar HVDC transmission line using different Back propagation Algorithms of ANN | |
| CN117491793B (en) | Hydrogen electric coupling system comprehensive performance test method, device and medium | |
| Wang et al. | A svm transformer fault diagnosis method based on improved bp neural network and multi-parameter optimization |
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 |