CN113495215B - Virtual battery cell system, operation method thereof and twin battery - Google Patents
Virtual battery cell system, operation method thereof and twin battery Download PDFInfo
- Publication number
- CN113495215B CN113495215B CN202110717551.XA CN202110717551A CN113495215B CN 113495215 B CN113495215 B CN 113495215B CN 202110717551 A CN202110717551 A CN 202110717551A CN 113495215 B CN113495215 B CN 113495215B
- Authority
- CN
- China
- Prior art keywords
- charge
- battery
- discharge
- battery cell
- virtual
- 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
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012549 training Methods 0.000 claims abstract description 36
- 238000004088 simulation Methods 0.000 claims abstract description 17
- 238000003062 neural network model Methods 0.000 claims abstract description 6
- 230000008569 process Effects 0.000 claims description 29
- 238000013528 artificial neural network Methods 0.000 claims description 28
- 230000003044 adaptive effect Effects 0.000 claims description 8
- 230000005284 excitation Effects 0.000 claims description 7
- 238000002474 experimental method Methods 0.000 claims description 7
- 238000007405 data analysis Methods 0.000 claims description 4
- 238000009795 derivation Methods 0.000 claims description 4
- 238000013500 data storage Methods 0.000 claims description 3
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 230000000087 stabilizing effect Effects 0.000 claims description 3
- 238000012360 testing method Methods 0.000 description 14
- 238000007599 discharging Methods 0.000 description 9
- 238000005259 measurement Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 238000012827 research and development Methods 0.000 description 5
- 238000007726 management method Methods 0.000 description 4
- 230000032683 aging Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 230000001276 controlling effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000002238 attenuated effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 230000001351 cycling effect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000011056 performance test Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- 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
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
- Secondary Cells (AREA)
Abstract
The invention discloses a virtual battery cell system, an operation method thereof and a twin battery, wherein the virtual battery cell system comprises an upper computer, a control unit and an execution unit, a data driving battery model is arranged in the upper computer, and the data driving battery model is obtained by inputting charge and discharge operation data of a physical battery cell into a neural network model for training; the upper computer is connected with the control unit to send the simulation parameters output by the data driving battery model to the control unit; the control unit judges the charge and discharge state according to the received simulation parameters of the data driving battery model to generate a charge and discharge control command, and is connected with the execution unit to send the charge and discharge control command to the execution unit; the execution unit outputs the charge-discharge characteristics according to the received charge-discharge control command, measures the electrical signals when the charge-discharge characteristics are output, and returns the electrical signals to the control unit. The invention can simulate the self-running characteristic of the physical battery cell with high precision.
Description
Technical Field
The invention relates to the technical field of virtual batteries, in particular to a virtual battery cell system, an operation method thereof and a twin battery.
Background
The battery is a core component of the electric automobile, and the Battery Management System (BMS) is a control center of the power battery of the electric automobile, is a core management system for monitoring and analyzing the state of the battery, controlling and managing the energy and collecting and transmitting information, and is a core for maintaining the normal operation of the whole battery. In the research and development and production process of the BMS, it is necessary to test the battery in different states to ensure the accuracy and safety thereof. However, the physical battery has a large capacity, and the discharging and charging processes are continuous processes, so that it is difficult to directly perform BMS test in a certain state of charge or aging state, and the time is long. Meanwhile, as the battery is repeatedly tested, the performance degradation of the battery may cause a decrease in accuracy, so that the performance test of the BMS is affected. Therefore, the virtual battery which can simulate the running characteristics of the physical battery in different states is researched and developed, the physical battery is replaced for testing in the research and development and production processes of the BMS, the testing time can be greatly shortened, the testing accuracy is improved, and the method has important values for improving the iteration speed of research and development of the BMS and reducing the research and development cost.
The Chinese patent document 201110192505.9 discloses a programmable control virtual battery module, which simulates the operation of a physical battery module through calculation and the virtual battery module connected with a computer so as to realize the charge-discharge response of the physical battery module and meet the requirements of BMS test. And the parameters of the virtual battery module are regulated through the control of the computer, and the virtual battery module simulates the electrical characteristic output of the physical battery module through a singlechip, a digital-to-analog converter, a direct current power supply and the like. Therefore, BMS testing can be performed instead of a physical battery module, safety problems in the testing process are eliminated, and the testing process is simplified.
The chinese patent document 201810700491.9 discloses a battery pack simulation system, which is configured by an upper computer, a plurality of middle computers and a plurality of groups of lower computers, and communicates with a BMS in real time to determine the operating state thereof. Through addding a plurality of median computers, carry out data acquisition to corresponding lower computer simultaneously, then gather and send to the host computer for the operating time of host computer, median computer and lower computer is synchronous, thereby has reduced operating time, has guaranteed battery operation dynamic curve's accuracy, so as to carry out BMS's test better.
There are many inventions in the field related to virtual batteries, but there are basically two more prominent problems: firstly, many virtual battery technologies do not have built-in battery models, so that corresponding electrical characteristics can be output only through a data set table look-up method, and the self-running state of the battery cannot be simulated. And the built-in battery models of part of technologies also adopt equivalent circuit models with limited precision, so that the running state and decay process of the built-in battery models cannot be simulated with high precision. Secondly, many virtual battery technologies can only simulate the discharging process of the battery, but cannot simulate the charging process of the battery, and the simulation characteristics are incomplete.
The foregoing background is only for the purpose of facilitating an understanding of the principles and concepts of the invention and is not necessarily in the prior art to the present application and is not intended to be used as an admission that such background is not entitled to antedate such novelty and creativity by the present application without undue evidence prior to the present application.
Disclosure of Invention
In order to solve the technical problems, the invention provides a virtual battery cell system, an operation method thereof and a twin battery, which can simulate the self-operation characteristic of a physical battery cell with high precision.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
the embodiment of the invention discloses a virtual battery cell system, which comprises an upper computer, a control unit and an execution unit, wherein a data driving battery model is built in the upper computer, and the data driving battery model is obtained by inputting charge and discharge operation data of a physical battery cell into a neural network model for training; the upper computer is connected with the control unit to send the simulation parameters output by the data driving battery model to the control unit; the control unit judges a charge and discharge state according to the received simulation parameters of the data-driven battery model to generate a charge and discharge control command, and is connected with the execution unit to send the charge and discharge control command to the execution unit; the execution unit outputs the charge-discharge characteristics according to the received charge-discharge control command, measures an electrical signal when the charge-discharge characteristics are output, and transmits the electrical signal back to the control unit.
Preferably, inputting charge and discharge operation data of the physical battery cell into the neural network model for training specifically includes: performing a multi-working-condition charge-discharge experiment based on a physical cell, collecting external characteristic data covering all voltage working intervals, training and verifying by adopting a neural network method based on the external characteristic data, and establishing a data-driven battery model; the multi-working-condition charge-discharge experiment is to fully charge the physical battery cell to 100% of charge state and then discharge the battery cell to cut-off voltage under different working conditions.
Preferably, the neural network model is an adaptive wavelet neural network.
Preferably, the training process of the adaptive wavelet neural network includes: the self-adaptive gradient descent method is adopted, forward propagation from input to output is carried out on each layer through the functions of the weight values and the excitation functions, then the weight values are updated in a chained derivation mode based on the error functions, and the next training is carried out after all the weight values are updated, the repeated circulation is carried out until the errors reach the training requirements, and the data-driven battery model training based on the self-adaptive wavelet neural network is completed.
Preferably, the upper computer communicates with the control unit in real time through an external interface of the upper computer, and the external interface of the upper computer is generated by software of the upper computer and is used for controlling a command function of the execution unit.
Preferably, the execution unit comprises a programmable power supply and a programmable load, wherein the programmable power supply is externally connected with an external power supply to form direct current output through a rectifying circuit and a voltage stabilizing circuit, and the programmable load is internally provided with a functional dissipation device.
Preferably, the execution unit and the control unit communicate in real time through a USB interface.
The other embodiment of the invention discloses an operation method of the virtual battery cell system, which is characterized by comprising the following steps:
s1: setting the initial state of the battery cell and selecting an operation mode in the upper computer, simulating the operation of the battery cell by the data-driven battery model, and outputting simulation parameters;
s2: the control unit receives the simulation parameters, judges the charge and discharge state to generate a charge and discharge control command, and sends the charge and discharge control command to the execution unit;
s3: and the execution unit outputs the charge and discharge characteristics according to the received charge and discharge control command, measures an electrical signal when the charge and discharge characteristics are output, and transmits the electrical signal back to the control unit and the upper computer to complete data analysis and storage.
Preferably, the performing unit performs the charge-discharge characteristic output according to the received charge-discharge control command specifically includes: the execution unit receives the charge and discharge control command and outputs electrical characteristics of the programmable power supply or the programmable load which form the execution unit, wherein the electrical characteristics of the output are voltage and current.
The invention also discloses a twin battery, which is formed by connecting multiple virtual battery core systems in series and/or in parallel.
Compared with the prior art, the invention has the beneficial effects that: according to the virtual battery cell system and the operation method thereof, the battery model is built for the highly nonlinear battery system based on the physical battery cell operation data through the strong nonlinear fitting capability of the neural network, and the battery model is built into the upper computer of the virtual battery cell system, so that the self-operation characteristic of the physical battery cell can be simulated with high precision. Meanwhile, a plurality of virtual battery core systems are combined in a certain serial-parallel mode to form a twin battery, the twin battery can mirror image and reflect the operation characteristics of a physical battery pack, even battery management is carried out based on wireless communication and cloud computing technology auxiliary BMS, the operation performance of the vehicle-mounted battery is improved, and the development direction of intelligent battery control is achieved.
In a further scheme, the invention can simulate the complete discharging and charging process of the battery through the output of the programmable power supply and the programmable load, and is closer to the actual running state of the physical battery cell.
Drawings
FIG. 1 is a block diagram of a virtual cell system in accordance with a preferred embodiment of the present invention;
FIG. 2 is a schematic diagram of the data-driven battery model creation process according to the preferred embodiment of the present invention;
FIG. 3 is a schematic diagram of an adaptive wavelet neural network training process according to a preferred embodiment of the present invention;
FIG. 4 is a flowchart of the operation of the virtual cell system of the preferred embodiment of the present invention;
FIGS. 5a and 5b are graphs comparing the operation effects of the virtual cell system and the existing equivalent circuit model according to the preferred embodiment of the present invention;
fig. 6 is a schematic diagram of a twin battery assisted mode of operation of a virtual cell system in accordance with a preferred embodiment of the present invention.
Detailed Description
In order to make the technical solution and advantages of the present invention more clear, the technical solution of the embodiments of the present invention will be fully described below with reference to the accompanying drawings in the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
The embodiment of the invention discloses a virtual battery cell system based on a data driving model, which comprises three modules, namely an upper computer, a control unit and an execution unit, wherein:
the upper computer is a virtual cell system control module, is internally provided with a data driving battery model, is a model obtained by training charge and discharge operation data of a certain type of cells through a neural network, can simulate charge and discharge characteristics of the cells with high precision, and has self-operation and capacity decay characteristics of the battery;
the control unit is an execution unit control module, and is communicated with the upper computer in real time through an external interface of the upper computer, receives simulation parameters of the battery model, simultaneously transmits a charge and discharge control command to the execution unit in real time through a communication port, and returns an actual output measured value of the execution unit; wherein the external interface of the upper computer is generated by upper computer software and can control the command function of the execution unit.
The execution unit is an execution module based on hardware equipment and comprises a programmable power supply and a programmable load, and the real-time communication with the control unit is used for completing the charge-discharge characteristic output of the virtual battery cell, measuring and returning an actual output electrical signal so as to be ready for carrying out data analysis. Wherein the real-time communication between the execution unit and the control unit is USB data transmission; the programmable power supply and the programmable load are provided with output interfaces, and have the same capacity of being connected with a physical battery cell and driving the load; the programmable power supply is externally connected with 220V household electricity, and forms direct current output through the rectifying circuit and the voltage stabilizing circuit, so that the power supply is an energy source in the virtual cell module discharging process; the programmable load is internally provided with a functional dissipation device which can consume the energy for charging the virtual battery cell.
The operation steps of the virtual battery cell system based on the data driving model comprise: s1: setting the initial state of the battery cell and selecting an operation mode by the upper computer, simulating the operation of the battery cell by the data-driven battery model, and outputting simulation parameters; s2: the control unit receives the simulation parameters, judges the charging or discharging state and sends a control command to the execution unit; s3: the execution unit receives the charge and discharge control command, completes the electrical characteristic output through the programmable power supply or the load, and simultaneously measures the output value and returns the output value to the control unit and the upper computer to complete the data analysis and storage; the output electrical characteristics are voltage and current, and the range of the output electrical characteristics is determined by the type of the battery cell of the data used for building the battery model.
The virtual battery cell system and the operation method thereof provided by the embodiment of the invention are characterized in that the operation model of the battery cell is obtained by training the physical battery cell data by adopting a neural network, and the operation model is built in the virtual battery cell to simulate the operation characteristics of the physical battery cell with high precision, so that the precision of the virtual battery cell is improved; and compared with the traditional equivalent circuit model, the data-driven battery model can better simulate the characteristics of capacity decay and the like in the operation process of the physical battery cell.
As shown in fig. 1, the system frame of an embodiment of the present invention is that the virtual battery cell based on the data driving battery model of the present invention includes three modules, namely, a host computer 10 with the data driving battery model built therein, a control unit 20 based on charge and discharge control and measurement feedback, and an execution unit 30 based on a programmable power supply and a load. The upper computer 10 mainly comprises an upper computer interface 11 and a built-in model 12, and the upper computer 10 is connected with the control unit 20 through an upper computer external interface 13. The control unit 20 includes two parts, namely a charge-discharge control 21 and a measurement feedback 22, the charge-discharge control 20 mainly determines the charge or discharge process according to the simulation parameters of the built-in model 12, and the measurement feedback 22 is used for receiving the output measurement value from the execution unit 30 to return to the upper computer 10. The execution unit 30 mainly comprises a programmable power supply 31 and a programmable load 32, and is in real-time communication with the control unit 20 through a data transmission interface 33, and is used as an execution element for discharging and charging, the discharging energy is sourced from an external power supply 40, and the consumption of the charging energy depends on a functional dissipation device of the programmable load 32.
Compared with other technologies in the field of virtual batteries, one of the biggest differences of the invention is that a high-precision battery model is built in, as shown in fig. 2, a data-driven battery model building process diagram of one embodiment of the invention is shown, firstly, a multi-working-condition charge-discharge experiment is carried out based on one type of battery core, wherein the charge-discharge experiment needs to be ensured to be full to 100% of charge state and then discharged to cut-off voltage under different working conditions, and external characteristic data covering all voltage working intervals is acquired. And then training and verifying by adopting a neural network method based on the external characteristic data of the battery, and establishing a data-driven battery model. Because the data-driven battery model is built by completely relying on the external characteristic data of the battery, the original errors in the equivalent process of the model are eliminated, the training data cover all voltage working intervals, and the original errors among different similar battery cells are eliminated, so that the built model has higher precision, and the capacity of the battery is continuously attenuated along with the operation of the model, is attenuated to a certain degree to stop discharging, and completely has the same charge-discharge process and electrical performance parameters as the physical battery cells.
Based on for more clear descriptionThe process of establishing the data-driven battery model by the neural network is shown in fig. 3, which is a training process of the data-driven battery model based on the adaptive wavelet neural network according to an embodiment, and the training process of the neural network in the data-driven battery model is illustrated by taking the adaptive wavelet neural network as an example. Wherein a three-layer fully-connected neural network is established, the node number of an input layer is K=4, and input variables are current (I), state of charge (SOC) and voltage (U) at the last moment t-1 ) And the current variation (delta I), the hidden layer is a single hidden layer, the node number is L=25, and the excitation function is a wavelet functionThe number of output layer nodes is M=1, and the excitation function is a linear function +.>The ideal output value is the terminal voltage (U).
In the training process, an adaptive gradient descent method is adopted, data is transmitted forward from input to output as shown by solid arrows in the figure, and training is carried out on each layer through the action of weights and excitation functions. In x k Input variable, ω, representing the kth input layer node of the neural network kl Is the weight between the kth input layer node and the ith hidden layer node, ω' lm O is the weight between the node of the first hidden layer and the node of the m output layer m Representing the output of the mth output layer node. Thus the input h of the first hidden layer node l Can be expressed as:
the output h 'of which can be obtained by implicit layer excitation function processing' l The method comprises the following steps:
the hidden layer output can be obtained by weight calculation and the node input p of the mth output layer m ：
Thus, neural network output o is available via output layer excitation function processing 1 The method comprises the following steps:
the output error function e with reference to the corresponding true value U is defined as:
and after the forward training is finished, updating and correcting the weight value by adopting a chain derivation mode based on the error function of the formula (6), solving by adopting a sequence from near to far from output as shown by a dotted arrow in the figure until the next training is carried out after all parameters are updated, and repeatedly cycling until the error reaches the training requirement, thereby finishing the training of the data-driven battery model based on the self-adaptive wavelet neural network. In the process of two adjacent training processes of t and t+1, omega is used kl The update procedure is as follows for example:
ω kl t+1 ＝ω kl t +Δω kl t (7)
Δω kl t the chain derivation process of (1) is as follows:
wherein omega kl t 、ω kl t+1 Respectively representing weights between a kth input layer node and a ith hidden layer node in t and t+1th training; e, e t Represents the error function at the t-th training, o m t Representing the output of the mth output layer node at the t-th training,representing the output of the first hidden layer node at the time of the t-th training.
In the embodiment based on the self-adaptive wavelet neural network training, 2000 times of the training process are performed through multi-working-condition electric core data, and the data driving battery model can simulate the charge and discharge performance of the battery with high precision. The data driving battery model, the programmable power supply and the load are integrated to form the virtual battery cell system of the embodiment, and the structure of the virtual battery cell system is shown in fig. 1, so that the operation of a physical battery cell is simulated.
As shown in fig. 4, a virtual cell workflow diagram of an embodiment of the present invention is shown, and the virtual cell operation process of the present invention is as follows:
s1: firstly, setting an initial state and an operation mode of the battery cell in the upper computer, wherein the initial state comprises an initial voltage, an initial charge state and an initial internal resistance. The virtual battery cell has the advantages that compared with the physical battery cell, the virtual battery cell can start to operate from any state of charge or aging state (the available internal resistance represents), and the battery operation time is greatly reduced;
s2: when the setting of the initial state and the running mode is completed, the virtual battery cell starts to run, and the battery model simulates analog signals of various electrical properties in the running process of the battery cell in real time and sends the analog signals to the control unit through an external interface;
s3: the control unit judges the charge and discharge state according to the analog signal, sends a control command to the execution unit, and simultaneously receives a measurement signal from the measurement module and returns the measurement signal to the upper computer so as to analyze the running state;
s4: the execution unit receives the control command, respectively controls the power supply and the load to discharge and charge the electric characteristic output, and measures the actual output signal value through the measurement module.
The virtual battery cell according to the embodiment of the invention completes the operation flow, and the operation state of the physical battery can be simulated through the operation flow, and battery experiments and the like can be carried out on the BMS. The test time and cost are saved, and the test precision is improved.
In order to compare the operation effects of the present invention with those of other virtual battery technologies, an example of the operation effects of one embodiment of the present invention is shown in fig. 5a and 5 b. The neural network adopts a self-adaptive wavelet neural network, and the equivalent circuit model adopts a least square method with forgetting factors to carry out segmentation parameter identification. As can be seen from the terminal voltage output curve shown in fig. 5a, the output precision of the battery model built by the neural network is far higher than that of the equivalent circuit model, and is very close to the operation true value of the physical battery, and the error magnitudes of the two models can be seen more intuitively from the terminal voltage error curve shown in fig. 5 b. And due to capacity decay, internal resistance change and the like in the discharging process of the battery, the equivalent circuit model needs to conduct parameter identification in real time, and is more loaded in the practical application process. The neural network-based data driven battery model is a preferred solution for building virtual batteries.
Along with the rapid development of communication technology and the gradual improvement of cloud computing capability, the remote control of the electric automobile becomes reality gradually. The twin battery obtains the operation data of the physical battery in real time, and corrects the model to ensure the mirror image relationship with the physical battery; and meanwhile, the operation state of the physical battery can be corrected. Thus, the remote auxiliary BMS can perform battery management, and the running performance of the battery is better ensured.
The physical battery pack 200 is composed of a plurality of physical battery cells connected in series and parallel, and the twin battery 110 is composed of a plurality of virtual battery cells based on the same series-parallel mode. Fig. 6 is a schematic diagram of a twin battery auxiliary operation mode based on a virtual battery cell according to the present invention, which is an application scenario of the virtual battery cell according to the present invention, wherein the twin battery 110 is formed by a plurality of virtual battery cells according to the present invention connected in series and parallel. During the operation of the twin battery, the data transmission module of the vehicle-mounted BMS uploads the operation data of the electric vehicle to the data storage center 120 of the twin battery in real time. The established battery model is a generalized model established based on the operation data of the battery cells, and the model is modified and specifically positioned by downloading the data of the specific battery from the data storage center, so that the twin battery 110 with the mirror image relationship with the physical battery pack 200 is established. Meanwhile, the twin battery platform 100 can analyze the running road condition of the electric automobile, calculate the optimal battery running mode under the road condition and compare with the current actual running state to obtain the correction parameters of the current running mode. The correction parameters are transmitted to the on-vehicle BMS via the on-line control center 130 of the twin battery, thereby controlling the corresponding power battery to reach the optimal operation state. The operation efficiency and performance of the power battery can be greatly improved by using the cloud power high-computing power in the mode, and the operation efficiency and performance of the power battery can be greatly improved by using the mode.
The preferred embodiment of the invention discloses a method for realizing a virtual battery cell based on a data driving model, wherein the virtual battery cell is a digital battery with the same electrical property and thermodynamic property as a physical battery cell, and is the research basis of a twin battery. The virtual battery cell implementation method comprises an upper computer with a built-in data driving battery model, a control unit based on charge and discharge control and measurement feedback, and an execution unit based on a programmable power supply and a load. The upper computer is a virtual battery cell system control module, a battery model based on a neural network is built in the upper computer, and the upper computer is a data driving model obtained by training the neural network based on physical battery cell data, so that the operation characteristics of the battery cell can be simulated with high precision. The control unit is an execution unit control module, can communicate with the upper computer in real time, judges and controls the charge and discharge states of the battery, and collects output signals of the execution unit. The execution unit is an execution module based on hardware equipment, can communicate with the control unit in real time, and is used for completing the output of the charge-discharge characteristics, measuring and returning the actual output electrical signals. The virtual battery core can simulate battery operation from random charge state or aging state through the simulation control hardware equipment of the data driving battery model, replaces the physical battery core to test BMS, is beneficial to improving the speed and precision of BMS test, can provide a foundation for research and development of twin batteries, and has important significance for development of electric automobiles.
The above is a description of the working principle, workflow and application prospect of the virtual battery cell implementation method based on the data driving battery model, and in the description of the present specification, the descriptions of the terms "one embodiment" and "example" and the like mean that the specific features, structures or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms are not necessarily aimed at being combined in a suitable manner in the opposite embodiments or examples.
The foregoing describes a method for implementing a virtual cell based on a data-driven battery model according to the present invention in detail, and further describes the invention with reference to specific embodiments, it should be pointed out that the description of the above embodiments is not intended to be limiting but is merely for aiding in understanding the core concept of the invention, and any modifications of the invention and alternatives equivalent to the present product, which are within the scope of protection of the claims of the present invention, will be apparent to those of ordinary skill in the art without departing from the principles of the invention.
Claims (9)
1. The virtual battery cell system is characterized by comprising an upper computer, a control unit and an execution unit, wherein a data driving battery model is built in the upper computer, and the data driving battery model is obtained by inputting charge and discharge operation data of a physical battery cell into a neural network model for training, so that the upper computer can simulate the charge and discharge characteristics of the battery cell and has the self-operation and capacity decay characteristics of a battery; the upper computer is connected with the control unit to send the simulation parameters output by the data driving battery model to the control unit; the control unit judges a charge and discharge state according to the received simulation parameters of the data-driven battery model to generate a charge and discharge control command, and is connected with the execution unit to send the charge and discharge control command to the execution unit; the execution unit outputs charge and discharge characteristics according to the received charge and discharge control command, measures an electrical signal when the charge and discharge characteristics are output, and transmits the electrical signal back to the control unit;
the method for training the physical battery cell comprises the following specific steps of: performing a multi-working-condition charge-discharge experiment based on a physical cell, collecting external characteristic data covering all voltage working intervals, training and verifying by adopting a neural network method based on the external characteristic data, and establishing a data-driven battery model; the multi-working-condition charge-discharge experiment is to fully charge the physical battery cell to 100% of charge state and then discharge the battery cell to cut-off voltage under different working conditions.
2. The virtual battery cell system of claim 1, wherein the neural network model is implemented using an adaptive wavelet neural network.
3. The virtual battery cell system of claim 2, wherein the training process of the adaptive wavelet neural network comprises: the self-adaptive gradient descent method is adopted, forward propagation from input to output is carried out on each layer through the functions of the weight values and the excitation functions, then the weight values are updated in a chained derivation mode based on the error functions, and the next training is carried out after all the weight values are updated, the repeated circulation is carried out until the errors reach the training requirements, and the data-driven battery model training based on the self-adaptive wavelet neural network is completed.
4. The virtual battery cell system of claim 1, wherein the host computer communicates with the control unit in real time via an external interface of the host computer, the external interface of the host computer being generated by software of the host computer for controlling command functions of the execution unit.
5. The virtual battery cell system of claim 1, wherein the execution unit comprises a programmable power supply and a programmable load, wherein the programmable power supply is externally connected with an external power supply to form a direct current output through a rectifying circuit and a voltage stabilizing circuit, and the programmable load is internally provided with a functional dissipation device.
6. The virtual battery cell system of claim 1, wherein the execution unit communicates with the control unit in real-time via a USB interface.
7. A method of operating a virtual cell system as claimed in any one of claims 1 to 6, comprising the steps of:
s1: setting the initial state of the battery cell and selecting an operation mode in the upper computer, simulating the operation of the battery cell by the data-driven battery model, and outputting simulation parameters;
s2: the control unit receives the simulation parameters, judges the charge and discharge state to generate a charge and discharge control command, and sends the charge and discharge control command to the execution unit;
s3: and the execution unit outputs the charge and discharge characteristics according to the received charge and discharge control command, measures an electrical signal when the charge and discharge characteristics are output, and transmits the electrical signal back to the control unit and the upper computer to complete data analysis and storage.
8. The method for operating a virtual battery cell system according to claim 7, wherein the executing unit performs the charge-discharge characteristic output according to the received charge-discharge control command specifically includes: the execution unit receives the charge and discharge control command and outputs electrical characteristics of the programmable power supply or the programmable load which form the execution unit, wherein the electrical characteristics of the output are voltage and current.
9. A twin battery, characterized in that it is constituted by a plurality of virtual cell systems according to any one of claims 1 to 6 in series and/or in parallel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110717551.XA CN113495215B (en) | 2021-06-28 | 2021-06-28 | Virtual battery cell system, operation method thereof and twin battery |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110717551.XA CN113495215B (en) | 2021-06-28 | 2021-06-28 | Virtual battery cell system, operation method thereof and twin battery |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113495215A CN113495215A (en) | 2021-10-12 |
| CN113495215B true CN113495215B (en) | 2024-01-12 |
Family
ID=77997632
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110717551.XA Active CN113495215B (en) | 2021-06-28 | 2021-06-28 | Virtual battery cell system, operation method thereof and twin battery |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113495215B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114935722B (en) * | 2022-05-30 | 2024-01-12 | 武汉理工大学 | Lithium battery side end collaborative management method based on digital twin |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6064180A (en) * | 1996-10-29 | 2000-05-16 | General Motors Corporation | Method and apparatus for determining battery state-of-charge using neural network architecture |
| CN101908657A (en) * | 2010-07-06 | 2010-12-08 | 奇瑞汽车股份有限公司 | Battery simulating system |
| CN102346204A (en) * | 2011-07-11 | 2012-02-08 | 毛广甫 | Programmable controlled virtual battery module |
| CN102411128A (en) * | 2011-07-25 | 2012-04-11 | 华北电力大学（保定） | Virtual battery management system and application method thereof |
| WO2014043888A1 (en) * | 2012-09-21 | 2014-03-27 | 华北电力大学 | Charging and discharging condition simulation system and method for electric vehicle power battery |
| WO2018153116A1 (en) * | 2017-02-21 | 2018-08-30 | 山东大学 | Fractional-order kibam battery model factoring in nonlinear capacity characteristics and parameter recognition method |
| CN110492186A (en) * | 2019-07-02 | 2019-11-22 | 北京航空航天大学 | A kind of management method of the power battery module based on cloud control technology |
| US20200074297A1 (en) * | 2018-09-03 | 2020-03-05 | Lg Electronics Inc. | Device including battery |
| CN111027165A (en) * | 2019-07-19 | 2020-04-17 | 北京航空航天大学 | Power battery pack management system and method based on digital twinning |
| CN112162125A (en) * | 2020-09-11 | 2021-01-01 | 东莞市明信技术有限公司 | Method for realizing simulation battery |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106772063B (en) * | 2016-11-21 | 2018-03-20 | 华中科技大学 | A kind of method and its device for monitoring charge states of lithium ion battery and health status |
-
2021
- 2021-06-28 CN CN202110717551.XA patent/CN113495215B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6064180A (en) * | 1996-10-29 | 2000-05-16 | General Motors Corporation | Method and apparatus for determining battery state-of-charge using neural network architecture |
| CN101908657A (en) * | 2010-07-06 | 2010-12-08 | 奇瑞汽车股份有限公司 | Battery simulating system |
| CN102346204A (en) * | 2011-07-11 | 2012-02-08 | 毛广甫 | Programmable controlled virtual battery module |
| CN102411128A (en) * | 2011-07-25 | 2012-04-11 | 华北电力大学（保定） | Virtual battery management system and application method thereof |
| WO2014043888A1 (en) * | 2012-09-21 | 2014-03-27 | 华北电力大学 | Charging and discharging condition simulation system and method for electric vehicle power battery |
| WO2018153116A1 (en) * | 2017-02-21 | 2018-08-30 | 山东大学 | Fractional-order kibam battery model factoring in nonlinear capacity characteristics and parameter recognition method |
| US20200074297A1 (en) * | 2018-09-03 | 2020-03-05 | Lg Electronics Inc. | Device including battery |
| CN110492186A (en) * | 2019-07-02 | 2019-11-22 | 北京航空航天大学 | A kind of management method of the power battery module based on cloud control technology |
| CN111027165A (en) * | 2019-07-19 | 2020-04-17 | 北京航空航天大学 | Power battery pack management system and method based on digital twinning |
| CN112162125A (en) * | 2020-09-11 | 2021-01-01 | 东莞市明信技术有限公司 | Method for realizing simulation battery |
Non-Patent Citations (4)
| Title |
|---|
| Experimental and Simulation Modal Analysis of a Prismatic Battery Module;Bizhong Xia 等;《ENERGIES》;全文 * |
| 基于虚拟电池技术的电动汽车充电设备测试系统;颜湘武等;《电力自动化设备》;第34卷(第11期);全文 * |
| 针对虚拟电池的HIL仿真平台研究;齐洪峰等;《电测与仪表》;第2页第1-2栏 * |
| 齐洪峰等.针对虚拟电池的HIL仿真平台研究.《电测与仪表》.2020,第2页第1-2栏. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113495215A (en) | 2021-10-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102411128B (en) | Virtual battery management system and application method thereof | |
| CN103926538B (en) | Change exponent number RC equivalent-circuit model based on AIC criterion and implementation method | |
| CN103632018B (en) | A kind of fuel cell modelling method based on Simscape platform | |
| CN101394096B (en) | Battery pack simulating device | |
| CN102169167B (en) | Device and method for detecting calculation accuracy of state of charge (SOC) of battery pack | |
| CN103293485A (en) | Model-based storage battery SOC (state of charge) estimating method | |
| CN111581850B (en) | Full-period power battery management system applying digital twinning technology | |
| Tang et al. | Design of power lithium battery management system based on digital twin | |
| Zhang et al. | Intelligent state of charge estimation of battery pack based on particle swarm optimization algorithm improved radical basis function neural network | |
| CN103675707A (en) | Method for evaluating lithium ion battery peak power online | |
| CN110303945B (en) | Self-adaptive optimization balance control method for electric quantity of storage battery pack | |
| Fan et al. | Simplified battery pack modeling considering inconsistency and evolution of current distribution | |
| CN105261795A (en) | Clustering algorithm-based method for bunching lithium-ion batteries in series | |
| CN106842056A (en) | One kind is based on two-step on-line intelligence optimized algorithm electrokinetic cell peak power method of estimation | |
| CN105116338B (en) | A kind of parallel connection type battery system modeling method based on SOC compensators | |
| Collet et al. | Multi-cell battery emulator for advanced battery management system benchmarking | |
| CN113495215B (en) | Virtual battery cell system, operation method thereof and twin battery | |
| CN110797933B (en) | Power battery active equalizing charging electrical control system and method | |
| CN204142849U (en) | For simulating battery simulation device and the system of charging electric vehicle | |
| CN107797011A (en) | Energy storage BMS balance policy analog detection method and device | |
| Sun et al. | The SOC estimation of NIMH battery pack for HEV based on BP neural network | |
| CN108427080A (en) | The state-of-charge computational methods of the power battery pack of hybrid power ship | |
| CN114492182A (en) | Method for predicting cycle life of battery pack in any topological structure | |
| Li et al. | Evaluation and analysis of circuit model for lithium batteries | |
| CN104730938A (en) | Battery characteristic simulation method and system |
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 |