CN111525197B - Storage battery SOH real-time estimation system and method - Google Patents
Storage battery SOH real-time estimation system and method Download PDFInfo
- Publication number
- CN111525197B CN111525197B CN202010362072.6A CN202010362072A CN111525197B CN 111525197 B CN111525197 B CN 111525197B CN 202010362072 A CN202010362072 A CN 202010362072A CN 111525197 B CN111525197 B CN 111525197B
- Authority
- CN
- China
- Prior art keywords
- battery
- storage battery
- real
- time
- internal resistance
- 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
Images
Classifications
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
-
- 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
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
- H01M10/00—Secondary cells; Manufacture thereof
- H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
- H01M10/4285—Testing apparatus
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- 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
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Electrochemistry (AREA)
- General Chemical & Material Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Secondary Cells (AREA)
Abstract
The invention relates to a storage battery SOH real-time estimation system and a storage battery SOH real-time estimation method, which are characterized in that: the lead-acid storage battery monitoring system comprises a lead-acid storage battery, a storage battery monitoring sensor assembly and a monitoring system assembly; the invention can predict the dynamic internal resistance value of the lead storage battery in real time without carrying out complete charge-discharge experiments, thereby prolonging the service life of the battery; the RBF neural network is constructed by adopting the dynamic internal resistance, the temperature and the electrolyte density of the battery, and compared with the traditional BP neural network, the RBF neural network has stronger input and output mapping functions, more accurate output results and real-time estimation without power failure maintenance.
Description
Technical Field
The invention relates to the field of lead-acid storage battery health state monitoring, in particular to a storage battery SOH real-time estimation system and method.
Background
The direct current power supply equipment of the power system largely selects valve-regulated lead acid storage batteries (VRLA). The valve-controlled lead-acid storage battery has high reliability and small maintenance workload, so that the storage battery is actually in a state of no maintenance for a long time on many occasions, particularly on the occasions such as an unattended transformer substation and the like, the basic requirement of the operation, maintenance and management of the storage battery is seriously deviated, and the storage battery is damaged greatly in advance of the nominal service life of the storage battery.
At present, domestic power supply and distribution stations generally adopt a storage battery terminal voltage monitoring mode, and the state representation capability of the storage battery by the voltage is very weak in practice; the traditional regulation requires a storage battery monitoring and maintenance mode which is too complicated and can be carried out only when power failure maintenance is carried out; the station storage battery is in a floating charge state for a long time, and the dynamic internal resistance Ri of the battery and the SOH of the storage battery cannot be calculated and estimated by adopting a common method through charge and discharge data monitoring.
Therefore, a GM (1,1) gray model is adopted to predict the dynamic internal resistance Ri, and the battery SOH is estimated by combining an RBF neural network; and selecting dynamic internal resistance Ri, real-time temperature Ti and battery fluid density Di as input, using battery actual capacity Q as output, and constructing an RBF neural network with an input layer node of 3, an implicit layer node of 8 and an output layer node of 1 to estimate the SOH of the battery in real time.
Disclosure of Invention
The invention aims to provide a storage battery SOH real-time estimation system, which can solve the problem that a storage battery monitoring and maintenance mode required by the traditional regulation is too complicated and is inconvenient to carry out only when power failure maintenance is carried out.
In order to solve the technical problems, the technical scheme of the invention is as follows: the utility model provides a storage battery SOH real-time estimation system which innovation point lies in: the lead-acid storage battery monitoring system comprises a lead-acid storage battery, a storage battery monitoring sensor assembly and a monitoring system assembly;
the storage battery monitoring sensor assembly comprises a voltage detection unit, a current detection unit, an electrolyte density measurement unit and a temperature sensor unit; one end of each of the voltage detection unit, the current detection unit, the electrolyte density measurement unit and the temperature sensor unit is connected to the lead-acid storage battery, and the other end of each of the voltage detection unit, the current detection unit, the electrolyte density measurement unit and the temperature sensor unit outputs signals to be collected and transmitted to the central data processing center through the CAN bus;
the monitoring system component comprises a central control unit, a central data processing center, a battery parameter database and a display terminal; the central control unit controls the electrolyte density measuring unit through a CAN bus and receives a signal fed back by the storage battery monitoring sensor assembly; the central data processing center comprises a lead storage battery dynamic internal resistance prediction module adopting a grey model and a battery actual capacity estimation module adopting a neural network; the central data processing center is connected to the central control unit, and the battery parameter database and the display terminal are both connected to the central data processing center.
Further, the electrolyte density measuring unit comprises a resonance sampling pipe and a sampling drainage unit; the sampling liquid discharge unit is controlled by the central control unit to drive the resonance sampling pipe to receive a feedback signal of the storage battery monitoring sensor assembly.
Further, a grey model of a lead-acid storage battery GM (1,1), a corresponding value of liquid resonance frequency and liquid density and historical monitoring data of the lead-acid storage battery are recorded in the battery parameter database.
A storage battery SOH real-time estimation method is characterized by comprising the following innovation points: the specific estimation method is as follows: the central data processing center receives the output signal of the central control unit, compares the output signal with the corresponding value of the resonance frequency and the liquid density in the battery parameter database to obtain the real-time density of the battery electrolyte, and inputs the real-time voltage signal and the real-time current signal into the battery parameter database; a GM (1,1) grey model is recorded in the battery parameter database, and the grey model is updated on line according to the detected voltage signal and current signal to predict the real-time internal resistance of the storage battery; and the central data processing center inputs the real-time voltage, the dynamic internal resistance predicted value, the real-time temperature and the battery liquid density into the trained neural network model, outputs the estimated actual capacity of the battery and calculates the SOH of the battery.
Further, the method for predicting the real-time internal resistance of the storage battery is based on grey model prediction and comprises the following steps: calculating to obtain dynamic internal resistance of the storage battery according to the output voltage signal and the current signal of the sensor group, and generating an original data group r(0)(t); performing GM (1,1) gray model modeling according to the original data group; the modeling method comprises the following steps:
A. the internal resistance value of the storage battery is coded according to the sequence of the equal interval time to form a time sequence r(0)(t) (t is 1, 2.. N), the prediction time is t, and the number of raw data used for modeling each time is N;
B. first read from the original time series, r(0)(t-N+1),...,r(0)(t) a total of N data values are assigned to the modeling sequence R(0)(i)(i＝1,2,...N)；
C. Modeling method for R according to GM (1,1) model(0)(i) Modeling is carried out, and a storage battery internal resistance predicted value at the moment of t + l, t +2 is obtained;
D. when the central data processing center DPC calculates the resistance value r of the storage battery at the moment of t +1(0)At (t +1), r is read again(0)(t-N+1),...,r(0)(t) A total of N data are assigned to the modeling sequence R(0)(i)(i＝1,2,...N)；
E. Reestablishing a GM (1,1) prediction model, and predicting the predicted values of t +2 and t + 3.. time;
and repeating the steps, so that a group of dynamic information GM (1,1) prediction models can be established, and the online prediction of the dynamic internal resistance of the storage battery is realized.
Further, the calculating battery SOH is based on a gray model and a neural network; the control system outputs real-time voltage, battery temperature and electrolyte density data, the central data processing center predicts real-time dynamic internal resistance of the battery as an input vector based on a gray model, and estimates real-time capacity Q of the storage battery; according to the formula:
and calculating to obtain the real-time SOH of the storage battery.
Further, the training process of the trained neural network model is as follows: step (1) charge-discharge stage: under the condition of constant temperature of an experimental environment, constant-current charging is carried out on storage batteries of the same type in a brand-new state until the storage batteries reach a cut-off voltage, then constant-voltage charging is carried out until the input current is lower than a rated value, and at the moment, the storage batteries can be determined to be in a full-power state; discharging the fully charged battery at a constant discharge rate until a cut-off voltage is reached, and calculating to obtain the total discharge capacity Q of the battery;
and (2) an aging cycle stage: fully charging and discharging the battery for n times to degrade the battery capacity to be less than 60% of the original total capacity Q; classifying the data into n groups of matrix spaces by using different environmental temperatures, and respectively training the matrix spaces into n groups of neural networks; performing k times of experiments at each environmental temperature to ensure that the capacity of the battery is degraded to be less than 60% of the original total capacity Q, respectively detecting the voltage, the temperature of the battery, the density of the electrolyte, calculating the dynamic internal resistance, and measuring 5 types of data in total of the actual total capacity Q of the battery, wherein m times of data measurement is performed during each charging and discharging to form a k 5 m matrix space;
step (3), RBF neural network training stage:
a. selecting an input vector and an output vector, namely, selecting voltage Ui, battery temperature Ti and electrolyte density Di, and calculating the obtained dynamic internal resistance Ri to form an input vector X ═ { Ui Ti Di Ri }, and taking total capacity Q as an output vector Y ═ Q };
b. construction of hidden layer neurons
The important part of the neural network model is the construction of hidden layer neurons, and the number of the neurons is determined by network self-adjustment of the RBF neural network in the learning process; the neuron state vector of the hidden layer is defined as z ═ { z ═ z1,z2,z3,...,zHAnd if yes, the state of the input vector x corresponding to the jth neuron of the hidden layer is:
function in the above formula
Is the radial basis function (i.e., RBF). The state of the hidden layer neuron corresponding to the output vector y is:
ω ═ ω in the above formula1，ω2，…，ωHThe logarithm is a weight value of the hidden layer neuron mapped on the output layer, | | | | represents a Euclidean norm, namely the space length of the multidimensional vector; the neural network adopts a Gaussian function as a radial basis function; namely:
wherein v isjIs the central vector of the jth neuron node of the hidden layer, and the dimension is the same as the input vector x; sigmajIs the base width of the jth neuron node.
c. Learning method design of neural network
An on-line construction method is adopted, and an Orthogonal Least Square (OLS) method is applied to correction on the basis.
The invention has the advantages that:
1) the invention can predict the dynamic internal resistance value of the lead storage battery in real time without carrying out complete charge-discharge experiments, thereby prolonging the service life of the battery; the RBF neural network is constructed by adopting the dynamic internal resistance, the temperature and the electrolyte density of the battery, and compared with the traditional BP neural network, the RBF neural network has stronger input and output mapping functions, more accurate output results and real-time estimation without power failure maintenance.
Drawings
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
FIG. 1 is an overall structure diagram of a real-time SOH estimation system of a lead-acid storage battery according to the invention.
FIG. 2 is a schematic diagram of an electrolyte density measuring cell according to the present invention.
Fig. 3 is a schematic diagram of the dynamic internal resistance prediction of the storage battery based on the gray model.
FIG. 4 is a schematic diagram of a gray model and RBF neural network-based battery SOH estimation system of the present invention.
FIG. 5 is a flow chart of the RBF neural network training of the battery capacity estimation system of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings or the orientations or positional relationships that the products of the present invention are conventionally placed in use, and are only used for convenience in describing the present invention and simplifying the description, but do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," "third," and the like are used solely to distinguish one from another and are not to be construed as indicating or implying relative importance.
Furthermore, the terms "horizontal", "vertical" and the like do not imply that the components are required to be absolutely horizontal or pendant, but rather may be slightly inclined. For example, "horizontal" merely means that the direction is more horizontal than "vertical" and does not mean that the structure must be perfectly horizontal, but may be slightly inclined.
In the description of the present invention, it should also be noted that, unless otherwise explicitly specified or limited, the terms "disposed," "mounted," "connected," and "connected" are to be construed broadly and may, for example, be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
A battery SOH real-time estimation system as shown in fig. 1 to 5, comprising a lead-acid battery 1, a battery monitoring sensor assembly 2 and a monitoring system assembly 3;
the storage battery monitoring sensor assembly 2 comprises a voltage detection unit 21, a current detection unit 22, an electrolyte density measurement unit 23 and a temperature sensor unit 24; one end of each of the voltage detection unit 21, the current detection unit 22, the electrolyte density measurement unit 23 and the temperature sensor unit 24 is connected to the lead-acid storage battery 1, and the other end outputs signals which are collected and transmitted to the central data processing center 32 through the CAN bus 4;
the monitoring system component 3 comprises a central control unit 31, a central data processing center 32, a battery parameter database 33 and a display terminal 34; the central control unit 31 controls the electrolyte density measuring unit through the CAN bus 4 and receives a signal fed back by the storage battery monitoring sensor assembly 2; the central data processing center 32 comprises a lead storage battery dynamic internal resistance prediction module adopting a grey model and a battery actual capacity estimation module adopting a neural network; the central data processing center 32 is connected to the central control unit 31, and the battery parameter database 33 and the display terminal 34 are connected to the central data processing center.
The electrolyte density measuring unit 23 comprises a resonance sampling pipe 231 and a sampling drainage unit 232; the sampling drainage unit 231 is controlled by the central control unit 31, and drives the resonance sampling pipe 231 to receive a feedback signal of the storage battery monitoring sensor assembly 2.
The battery parameter database 33 records a gray model of the lead-acid storage battery GM (1,1), corresponding values of the liquid resonance frequency and the liquid density, and historical monitoring data of the lead-acid storage battery.
A storage battery SOH real-time estimation method comprises the following specific steps: the central data processing center receives the output signal of the central control unit, compares the output signal with the corresponding value of the resonance frequency and the liquid density in the battery parameter database to obtain the real-time density of the battery electrolyte, and inputs the real-time voltage signal and the real-time current signal into the battery parameter database; a GM (1,1) grey model is recorded in the battery parameter database, and the grey model is updated on line according to the detected voltage signal and current signal to predict the real-time internal resistance of the storage battery; and the central data processing center inputs the real-time voltage, the dynamic internal resistance predicted value, the real-time temperature and the battery liquid density into the trained neural network model, outputs the estimated actual capacity of the battery and calculates the SOH of the battery.
The method for predicting the real-time internal resistance of the storage battery is based on grey model prediction and comprises the following steps: calculating to obtain dynamic internal resistance of the storage battery according to the output voltage signal and the current signal of the sensor group, and generating an original data group r(0)(t); performing GM (1,1) gray model modeling according to the original data group; the modeling method comprises the following steps:
A. the internal resistance value of the storage battery is coded according to the sequence of the equal interval time to form a time sequence r(0)(t) (t is 1, 2.. N), the prediction time is t, and the number of raw data used for modeling each time is N;
B. first read from the original time series, r(0)(t-N+1),...,r(0)(t) a total of N data values are assigned to the modeling sequence R(0)(i)(i＝1,2,...N)；
C. Modeling method for R according to GM (1,1) model(0)(i) Modeling is carried out, and a storage battery internal resistance predicted value at the moment of t + l, t +2 is obtained;
D. when the central data processing center DPC calculates the resistance value r of the storage battery at the moment of t +1(0)At (t +1), r is read again(0)(t-N+1),...,r(0)(t) A total of N data are assigned to the modeling sequence R(0)(i)(i＝1,2,...N)；
E. Reestablishing a GM (1,1) prediction model, and predicting the predicted values of t +2 and t + 3.. time;
and repeating the steps, so that a group of dynamic information GM (1,1) prediction models can be established, and the online prediction of the dynamic internal resistance of the storage battery is realized.
Calculating battery SOH is based on a gray modelAnd a neural network; the control system outputs real-time voltage, battery temperature and electrolyte density data, the central data processing center predicts real-time dynamic internal resistance of the battery as an input vector based on a gray model, and estimates real-time capacity Q of the storage battery; according to the formula:
and calculating to obtain the real-time SOH of the storage battery.
The training process of the trained neural network model is as follows:
step (1) charge-discharge stage: under the condition of constant temperature of an experimental environment, constant-current charging is carried out on storage batteries of the same type in a brand-new state until the storage batteries reach a cut-off voltage, then constant-voltage charging is carried out until the input current is lower than a rated value, and at the moment, the storage batteries can be determined to be in a full-power state; discharging the fully charged battery at a constant discharge rate until a cut-off voltage is reached, and calculating to obtain the total discharge capacity Q of the battery;
and (2) an aging cycle stage: fully charging and discharging the battery for n times to degrade the battery capacity to be less than 60% of the original total capacity Q; classifying the data into n groups of matrix spaces by using different environmental temperatures, and respectively training the matrix spaces into n groups of neural networks; performing k times of experiments at each environmental temperature to ensure that the capacity of the battery is degraded to be less than 60% of the original total capacity Q, respectively detecting the voltage, the temperature of the battery, the density of the electrolyte, calculating the dynamic internal resistance, and measuring 5 types of data in total of the actual total capacity Q of the battery, wherein m times of data measurement is performed during each charging and discharging to form a k 5 m matrix space;
step (3), RBF neural network training stage:
a. input vector and output vector selection
Calculating the dynamic internal resistance Ri to obtain an input vector X ═ Ui Ti Di Ri and a total capacity Q as an output vector Y ═ Q by using the voltage Ui, the battery temperature Ti and the electrolyte density Di;
b. construction of hidden layer neurons
The important part of the neural network model is the construction of hidden layer neurons, and the number of the neurons is determined by network self-adjustment of the RBF neural network in the learning process; the neuron state vector of the hidden layer is defined as z ═{z1,z2,z3,...,zHAnd if yes, the state of the input vector x corresponding to the jth neuron of the hidden layer is:
function in the above formula
Is the radial basis function (i.e., RBF). The state of the hidden layer neuron corresponding to the output vector y is:
ω ═ ω in the above formula1，ω2，…，ωHThe logarithm is a weight value of the hidden layer neuron mapped on the output layer, | | | | represents a Euclidean norm, namely the space length of the multidimensional vector; the neural network adopts a Gaussian function as a radial basis function; namely:
wherein v isjIs the central vector of the jth neuron node of the hidden layer, and the dimension is the same as the input vector x; sigmajIs the base width of the jth neuron node.
c. Learning method design of neural network
An on-line construction method is adopted, and an Orthogonal Least Square (OLS) method is applied to correction on the basis.
It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (7)
1. A storage battery SOH real-time estimation system is characterized in that: the lead-acid storage battery monitoring system comprises a lead-acid storage battery, a storage battery monitoring sensor assembly and a monitoring system assembly;
the storage battery monitoring sensor assembly comprises a voltage detection unit, a current detection unit, an electrolyte density measurement unit and a temperature sensor unit; one end of each of the voltage detection unit, the current detection unit, the electrolyte density measurement unit and the temperature sensor unit is connected to the lead-acid storage battery, and the other end of each of the voltage detection unit, the current detection unit, the electrolyte density measurement unit and the temperature sensor unit outputs signals to be collected and transmitted to the central data processing center through the CAN bus;
the monitoring system component comprises a central control unit, a central data processing center, a battery parameter database and a display terminal; the central control unit controls the electrolyte density measuring unit through a CAN bus and receives a signal fed back by the storage battery monitoring sensor assembly; the central data processing center comprises a lead storage battery dynamic internal resistance prediction module adopting a grey model and a battery actual capacity estimation module adopting a neural network; the central data processing center is connected to the central control unit, and the battery parameter database and the display terminal are both connected to the central data processing center;
the estimation system adopts dynamic internal resistance, temperature and electrolyte density of a battery to construct an RBF neural network;
the real-time internal resistance of the battery is predicted based on a grey model.
2. The system of claim 1 for real-time estimation of SOH of a battery, wherein: the electrolyte density measuring unit comprises a resonance sampling pipe and a sampling and liquid discharging unit; the sampling liquid discharge unit is controlled by the central control unit to drive the resonance sampling pipe to receive a feedback signal of the storage battery monitoring sensor assembly.
3. The system of claim 1 for real-time estimation of SOH of a battery, wherein: and the battery parameter database records a gray model of the lead-acid storage battery GM (1,1), corresponding values of liquid resonance frequency and liquid density and historical monitoring data of the lead-acid storage battery.
4. A storage battery SOH real-time estimation method is characterized by comprising the following steps: the specific estimation method is as follows: the central data processing center receives the output signal of the central control unit, compares the output signal with the corresponding value of the resonance frequency and the liquid density in the battery parameter database to obtain the real-time density of the battery electrolyte, and inputs the real-time voltage signal and the real-time current signal into the battery parameter database; a GM (1,1) grey model is recorded in the battery parameter database, and the grey model is updated on line according to the detected voltage signal and current signal to predict the real-time internal resistance of the storage battery; and the central data processing center inputs the real-time voltage, the dynamic internal resistance predicted value, the real-time temperature and the battery liquid density into the trained neural network model, outputs the estimated actual capacity of the battery and calculates the SOH of the battery.
5. The method according to claim 4, wherein the method comprises the following steps: the method for predicting the real-time internal resistance of the storage battery is based on grey model prediction and comprises the following steps: calculating to obtain dynamic internal resistance of the storage battery according to the output voltage signal and the current signal of the sensor group, and generating an original data group r(0)(t); performing GM (1,1) gray model modeling according to the original data group; the modeling method comprises the following steps:
A. the internal resistance value of the storage battery is coded according to the sequence of the equal interval time to form a time sequence r(0)(t) (t is 1, 2.. N), the prediction time is t, and the number of raw data used for modeling each time is N;
B. first read from the original time series, r(0)(t-N+1),...,r(0)(t) a total of N data values are assigned to the modeling sequence R(0)(i)(i＝1,2,...N)；
C. Modeling method for R according to GM (1,1) model(0)(i) Modeling is carried out, and a storage battery internal resistance predicted value at the moment of t + l, t +2 is obtained;
D. when the central data processing center DPC calculates t +1Carved battery resistance r(0)At (t +1), r is read again(0)(t-N+1),...,r(0)(t) A total of N data are assigned to the modeling sequence R(0)(i)(i＝1,2,...N)；
E. Reestablishing a GM (1,1) prediction model, and predicting the predicted values of t +2 and t + 3.. time;
and repeating the steps, so that a group of dynamic information GM (1,1) prediction models can be established, and the online prediction of the dynamic internal resistance of the storage battery is realized.
6. The method according to claim 4, wherein the method comprises the following steps: the calculated battery SOH is based on a grey model and a neural network; the control system outputs real-time voltage, battery temperature and electrolyte density data, the central data processing center predicts real-time dynamic internal resistance of the battery as an input vector based on a gray model, and estimates real-time capacity Q of the storage battery; according to the formula:
and calculating to obtain the real-time SOH of the storage battery.
7. The method according to claim 4, wherein the method comprises the following steps: the training process of the trained neural network model is as follows:
step (1) charge-discharge stage: under the condition of constant temperature of an experimental environment, constant-current charging is carried out on storage batteries of the same type in a brand-new state until the storage batteries reach a cut-off voltage, then constant-voltage charging is carried out until the input current is lower than a rated value, and at the moment, the storage batteries can be determined to be in a full-power state; discharging the fully charged battery at a constant discharge rate until a cut-off voltage is reached, and calculating to obtain the total discharge capacity Q of the battery;
and (2) an aging cycle stage: fully charging and discharging the battery for n times to degrade the battery capacity to be less than 60% of the original total capacity Q; classifying the data into n groups of matrix spaces by using different environmental temperatures, and respectively training the matrix spaces into n groups of neural networks; performing k times of experiments at each environmental temperature to ensure that the capacity of the battery is degraded to be less than 60% of the original total capacity Q, respectively detecting the voltage, the temperature of the battery, the density of the electrolyte, calculating the dynamic internal resistance, and measuring 5 types of data in total of the actual total capacity Q of the battery, wherein m times of data measurement is performed during each charging and discharging to form a k 5 m matrix space;
step (3), RBF neural network training stage:
a. input vector and output vector selection
Calculating the dynamic internal resistance Ri to obtain an input vector X ═ Ui Ti Di Ri and a total capacity Q as an output vector Y ═ Q by using the voltage Ui, the battery temperature Ti and the electrolyte density Di;
b. construction of hidden layer neurons
The important part of the neural network model is the construction of hidden layer neurons, and the number of the neurons is determined by network self-adjustment of the RBF neural network in the learning process; the neuron state vector of the hidden layer is defined as z ═ { z ═ z1,z2,z3,...,zHAnd if yes, the state of the input vector x corresponding to the jth neuron of the hidden layer is:
function in the above formula
For the radial basis function (i.e., RBF), the state of the hidden layer neuron corresponding to the output vector y is:
ω ═ ω in the above formula1，ω2，...ωHThe logarithm is a weight value of the hidden layer neuron mapped on the output layer, | | | | represents a Euclidean norm, namely the space length of the multidimensional vector; the neural network adopts a Gaussian function as a radial basis function; namely:
wherein v isjIs the central vector of the jth neuron node of the hidden layer, and the dimension is the same as the input vector x; sigmajIs as followsA base width of j neuron nodes;
c. learning method design of neural network
An on-line construction method is adopted, and an Orthogonal Least Square (OLS) method is applied to correction on the basis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010362072.6A CN111525197B (en) | 2020-04-30 | 2020-04-30 | Storage battery SOH real-time estimation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010362072.6A CN111525197B (en) | 2020-04-30 | 2020-04-30 | Storage battery SOH real-time estimation system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111525197A CN111525197A (en) | 2020-08-11 |
| CN111525197B true CN111525197B (en) | 2021-07-30 |
Family
ID=71912129
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010362072.6A Active CN111525197B (en) | 2020-04-30 | 2020-04-30 | Storage battery SOH real-time estimation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111525197B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113296009B (en) * | 2021-04-23 | 2023-03-14 | 重庆大学 | Method for predicting and recombining residual life of retired power lithium ion battery |
| CN114186756A (en) * | 2022-01-11 | 2022-03-15 | 国网湖北省电力有限公司孝感供电公司 | Method for predicting energy storage capacity of storage battery of power distribution network terminal |
Family Cites Families (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3935099B2 (en) * | 2003-04-15 | 2007-06-20 | 株式会社デンソー | Internal state detection system for power storage device for vehicle |
| CN202693771U (en) * | 2012-05-11 | 2013-01-23 | 山东轻工业学院 | Portable storage battery state detection system |
| CN102680903B (en) * | 2012-05-11 | 2015-01-28 | 齐鲁工业大学 | Portable storage battery state detection system and method |
| CN103728569B (en) * | 2014-01-09 | 2016-03-09 | 中国计量学院 | A kind of accumulator capacity Forecasting Methodology based on multifactor Grey Relational Model |
| CN103983919B (en) * | 2014-05-28 | 2016-10-05 | 山东大学 | A kind of based on GM(1, N) the battery life predicting method of gray model |
| CN106033113B (en) * | 2015-03-19 | 2019-03-08 | 国家电网公司 | A kind of energy-storage battery group health state evaluation method |
| CN105807230B (en) * | 2016-03-11 | 2019-03-26 | 郑贵林 | Remaining battery capacity and health status rapid detection method and device |
| CN205941261U (en) * | 2016-08-28 | 2017-02-08 | 常州市鑫祥科尔威船舶设备制造有限公司 | Liquid density measuring apparatu |
| CN207938741U (en) * | 2018-03-26 | 2018-10-02 | 厦门忻德信息技术有限公司 | A kind of lead-acid power accumulator monitoring device |
| CN109782176A (en) * | 2018-12-20 | 2019-05-21 | 上海交通大学 | Battery remaining power On-line Estimation method based on NARX model |
| CN110058159A (en) * | 2019-04-29 | 2019-07-26 | 杭州电子科技大学 | A kind of lithium battery health status estimation method based on grey neural network |
| CN110556597B (en) * | 2019-09-03 | 2020-11-24 | 浙江大学 | Operation and maintenance system and method for valve-controlled sealed lead-acid storage battery |
| CN110632528B (en) * | 2019-11-04 | 2021-08-31 | 桂林电子科技大学 | Lithium battery SOH estimation method based on internal resistance detection |
| CN110703121A (en) * | 2019-11-08 | 2020-01-17 | 北京化工大学 | Lithium ion battery health state prediction method |
-
2020
- 2020-04-30 CN CN202010362072.6A patent/CN111525197B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111525197A (en) | 2020-08-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102324582B (en) | Intelligent maintenance device of multifunctional lead-acid battery and capacity prediction method | |
| CN107390127A (en) | A kind of SOC estimation method | |
| CN114372417A (en) | Electric vehicle battery health state and remaining life evaluation method based on charging network | |
| CN205693405U (en) | Set of cells or capacitor group management system | |
| CN110658462B (en) | Lithium battery online service life prediction method based on data fusion and ARIMA model | |
| CN111525197B (en) | Storage battery SOH real-time estimation system and method | |
| CN111579993B (en) | Lithium battery capacity online estimation method based on convolutional neural network | |
| Pei et al. | An Equivalent Circuit Model for Lithium Battery of Electric Vehicle considering Self‐Healing Characteristic | |
| CN110658459A (en) | Lithium ion battery state of charge estimation method based on bidirectional cyclic neural network | |
| CN114035098A (en) | Lithium battery health state prediction method integrating future working condition information and historical state information | |
| CN112881914A (en) | Lithium battery health state prediction method | |
| US20230213587A1 (en) | Method and System for Efficiently Monitoring Battery Cells of a Device Battery in an External Central Processing Unit Using a Digital Twin | |
| CN115219932A (en) | Method and device for evaluating the relative aging state of a battery of a device | |
| CN115621573A (en) | System and method for predicting health state of lithium battery of energy storage power station | |
| CN117828408B (en) | Energy storage capacity data processing method and system | |
| CN114781875A (en) | Micro-grid economic operation state evaluation method based on deep convolutional network | |
| CN115101836A (en) | Cloud-end converged battery system management method and device | |
| Li et al. | Research on state-of-charge Estimation of Lithium-ion Batteries Based on Improved Sparrow Search Algorithm-BP Neural Network | |
| CN112327169B (en) | Lithium battery residual life prediction method | |
| CN110232432B (en) | Lithium battery pack SOC prediction method based on artificial life model | |
| CN116736141A (en) | Lithium battery energy storage safety management system and method | |
| CN117406092A (en) | State estimation method, device and system for distributed modular battery energy storage system | |
| CN113098084A (en) | Battery Internet of things system | |
| CN116500475A (en) | Energy storage acquisition method and system with real-time SOC correction compensation | |
| CN115656837A (en) | Fault prediction method for series-connected battery |
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 |