CN111948563B - Electric forklift lithium battery residual life prediction method based on multi-neural network coupling - Google Patents

Electric forklift lithium battery residual life prediction method based on multi-neural network coupling Download PDF

Info

Publication number
CN111948563B
CN111948563B CN202010568146.1A CN202010568146A CN111948563B CN 111948563 B CN111948563 B CN 111948563B CN 202010568146 A CN202010568146 A CN 202010568146A CN 111948563 B CN111948563 B CN 111948563B
Authority
CN
China
Prior art keywords
lithium battery
time
prediction
input
open
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
Application number
CN202010568146.1A
Other languages
Chinese (zh)
Other versions
CN111948563A (en
Inventor
童水光
童哲铭
李元松
苗嘉智
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University ZJU
Hangcha Group Co Ltd
Original Assignee
Zhejiang University ZJU
Hangcha Group Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhejiang University ZJU, Hangcha Group Co Ltd filed Critical Zhejiang University ZJU
Priority to CN202010568146.1A priority Critical patent/CN111948563B/en
Publication of CN111948563A publication Critical patent/CN111948563A/en
Application granted granted Critical
Publication of CN111948563B publication Critical patent/CN111948563B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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]
    • G01R31/396—Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery
    • 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
    • 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]
    • G01R31/392—Determining battery ageing or deterioration, e.g. state of health
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Abstract

The invention provides a method for predicting the remaining life of a lithium battery of an electric forklift based on multi-neural-network coupling, aiming at overcoming the problems of complex calculation, long time consumption and low prediction precision in the process of predicting the remaining life of the lithium battery in the prior art, improving the prediction calculation precision and reducing the training time of a prediction model, and comprising the following steps of: establishing an open-circuit voltage prediction model based on a long-time and short-time memory neural network, and optimizing the network by adopting an RMSprop algorithm and a dropout regularization method, thereby predicting the open-circuit voltage value V of the lithium battery in a discharge cycleOC(ii) a Dividing the prediction result into a plurality of discharge cycles in sequence, and counting the number N of open-circuit voltage samples from the initial voltage to the minimum voltage in each discharge cycleSUsing the sampling time TSThe same, the time T to discharge to the minimum voltage in each discharge cycle is obtainedmin(ii) a And establishing a capacity prediction model based on an artificial neural network to predict the capacity C of the lithium battery so as to obtain a predicted value RUL of the residual life of the lithium battery.

Description

Electric forklift lithium battery residual life prediction method based on multi-neural network coupling
Technical Field
The invention relates to the technical field of batteries, in particular to a method for predicting the remaining life of a battery based on multi-neural-network coupling.
Background
The lithium battery has high energy density, long cycle life and high discharge platform, and can provide reliable power sources for various engineering machines. At present, the lithium battery is adopted by the electric forklift as a power source, the electrochemical performance of the lithium battery can be gradually reduced under the condition of continuous charging and discharging, the capacity is slowly attenuated, the attenuation speed of the performance of the lithium battery is accelerated after the maximum service life is reached, the heat productivity is increased, the safety is reduced, the equipment is easy to operate and unstable, economic loss is caused, and safety accidents can be caused seriously or even. Therefore, the lithium battery needs to be maintained and replaced in time, and the attenuation condition of the battery capacity can be given in advance by the residual life prediction method, so that a reasonable maintenance scheme can be made in advance, and the stable operation of the electric forklift is ensured. Residual life prediction refers to predicting the number of cycles that a battery can still run before the end of life is reached, given the partial capacity fade data.
The traditional residual life prediction method is a model-based method, which establishes a model based on a battery failure principle and an electrochemical reaction and predicts the battery life attenuation condition, and has the defects of complex calculation and low precision.
Disclosure of Invention
The invention provides a method for predicting the residual life of a lithium battery of an electric forklift based on multi-neural-network coupling, aiming at solving the problems of complex calculation, long time consumption and low prediction precision in the process of predicting the residual life of the lithium battery in the prior art, so that the prediction calculation precision is improved, and the training time of a prediction model is reduced.
In order to achieve the purpose, the invention adopts the following technical scheme, which comprises the following steps:
s1: establishing an open-circuit voltage prediction model based on a long-time and short-time memory neural network, and optimizing the network by adopting an RMSprop algorithm and a dropout regularization method, thereby predicting the open-circuit voltage value V of the lithium battery in a discharge cycleOC；
S2: dividing the prediction result into a plurality of discharge cycles in sequence, and counting the number N of open-circuit voltage samples from the initial voltage to the minimum voltage in each discharge cycleSUsing the sampling time TSThe same, the time T to discharge to the minimum voltage in each discharge cycle is obtainedmin；
S3: and establishing a capacity prediction model based on an artificial neural network to predict the capacity C of the lithium battery so as to obtain a predicted value RUL of the residual life of the lithium battery.
The long-time and short-time memory neural network and the artificial neural network are combined to establish the residual life prediction model of the lithium battery of the electric forklift, the fitting capacity and the prediction capacity of the long-time and short-time memory neural network to nonlinear data are fully utilized, the advantage of simple structure of the artificial neural network is utilized, the prediction model is optimized by the RMSprop algorithm and the dropout regularization method, the model calculation precision is improved, and the model training time is shortened.
Preferably, the specific steps of S1 are as follows:
s11: setting the number of input neurons to NI1Setting the number of the nerve hiding layers to be LH1And number of hidden layer neurons NH1Establishing a relation mapping between input parameters and output parameters by using a sliding window, and establishing an input V during training if the length of the sliding window is L and the length of training data is MinputTo outputs of V respectivelyoutputThe mapping relation of (2) is shown in formula (1):
Figure BDA0002548270080000021
wherein, VinputEach row in the set corresponds to an input sample with a length of L;
s12: model obtained by training S11 is utilized to split open-circuit voltage VOCThe prediction is carried out by adopting the following method, wherein the input and the output are shown as the formula (2):
Figure BDA0002548270080000031
wherein, VOCEach row in the circuit diagram corresponds to a predicted input sample, and each prediction obtains a predicted value of the open-circuit voltage
Figure BDA0002548270080000032
Using the obtained predicted values
Figure BDA0002548270080000033
Updating the input sample sequence for the next prediction, i.e. the predicted value
Figure BDA0002548270080000034
Adding the data to the end of the input sample sequence, and deleting the first data of the input sample sequence, thereby completing the window sliding operation and obtaining all predicted values of the open-circuit voltage
Figure BDA0002548270080000035
Preferably, the specific steps of S2 are as follows:
s21: the method for sequentially dividing the prediction result into a plurality of discharge cycles is as follows:
if the time TjSo that equation (15) holds:
Figure BDA0002548270080000036
then Tj(j＝0，1，2，…，n；T01) is the end time of the jth cycle, and the open circuit voltage sequence of the jth cycle is
Figure BDA0002548270080000037
S22: counting the number N of open-circuit voltage samples from the initial voltage to the minimum voltage in each discharge cycleSUsing the sampling time TSThe same, the time T to discharge to the minimum voltage in each discharge cycle is obtainedminAs shown in formula (16):
Tmin＝NS×TS (16)
preferably, the specific steps of S3 are as follows:
s31: setting the number of input neurons to NI2Setting the number of the nerve hiding layers to be LH2And number of hidden layer neurons NH2Establishing the time T from discharge to minimum voltage in the discharge cycleminMapping relation to lithium battery capacity C;
s32: the calculation process of each neuron of the hidden layer is shown as the formula (17):
HT＝σ(WHTTmin+bH)； (17)
HTrepresenting a hidden layer computation function, where WHTAnd bHInputting a weight matrix and a bias parameter matrix of the function input value; after 20 times of iteration, the training is finished, and the model is saved. The predicted time T for discharging to the minimum voltage in each discharge cycleminInputting a model to obtain a capacity prediction result;
s33: defining EOL as capacity CjEqual to rated capacity C0The number j of cycles corresponding to 70%; defining and calculating the residual life RUL of the lithium battery as shown in formula (18):
RUL＝UL-EOL； (18)
wherein, UL is the number of the operating cycles of the current lithium battery, and EOL is the end point of the service life of the lithium battery.
The invention has the advantages of
The invention establishes a multi-neural-network coupling model of the long-time memory-artificial neural network, and effectively improves the prediction precision. The RMSprop algorithm is used for training, so that the accuracy of model prediction is improved; the model is optimized by using the dropout method, the overfitting problem is avoided, the model training time is reduced, and the model prediction accuracy is improved. The open-circuit voltage is used as prediction input data, and compared with a long-time memory-artificial neural network coupling model and a simple long-time memory neural network model which adopt the discharge time as characteristic parameters, the accuracy of a prediction result is obviously improved, and the residual life of the lithium battery of the electric forklift can be accurately predicted.
Drawings
FIG. 1 is a general flow diagram of the present invention;
FIG. 2 is an analysis of a first set of predictors according to the present invention;
FIG. 3 is an analysis of a second set of predictors according to the present invention;
FIG. 4 is an absolute error of remaining life prediction;
figure 5 is the root mean square error of the capacity fade prediction.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
The invention discloses a multi-neural-network-coupling-based method for predicting the remaining life of a lithium battery of an electric forklift, which is shown in a general flow chart in figure 1, and comprises the following specific operation processes:
s1: establishing an open-circuit voltage prediction model based on a long-time and short-time memory neural network, and optimizing the network by adopting an RMSprop algorithm and a dropout regularization method, thereby predicting the open-circuit voltage value V of the lithium battery in a discharge cycleOCThe method comprises the following specific steps:
s11: setting the number of input neurons to NI1Setting the number of the nerve hiding layers to be LH1And number of hidden layer neurons NH1Establishing a relation mapping between input parameters and output parameters by using a sliding window, and establishing an input V during training if the length of the sliding window is L and the length of training data is MinputTo outputs of V respectivelyoutputThe mapping relation of (2) is shown in formula (1):
Figure BDA0002548270080000051
wherein, VinputEach row in the set corresponds to an input sample with a length of L;
s12: model obtained by training S11 is utilized to split open-circuit voltage VOCThe prediction is carried out by adopting the following method, wherein the input and the output are shown as the formula (2):
Figure BDA0002548270080000052
wherein, VOCEach row in the circuit diagram corresponds to a predicted input sample, and each prediction obtains a predicted value of the open-circuit voltage
Figure BDA0002548270080000053
Using the obtained predicted values
Figure BDA0002548270080000054
For the next timeUpdating the predicted input sample sequence, i.e. predicting the value
Figure BDA0002548270080000055
Adding the data to the end of the input sample sequence, and deleting the first data of the input sample sequence, thereby completing the window sliding operation and obtaining all predicted values of the open-circuit voltage
Figure BDA0002548270080000061
The long-time and short-time memory neural network has good prediction and fitting capacity for time series data, in order to prevent overfitting, dropout regularization processing is carried out on the network, namely vectors of 0 and 1 are randomly generated, so that part of neurons in the network are randomly discarded, and the formulas (3) and (4) are shown as follows:
Figure BDA0002548270080000062
Figure BDA0002548270080000063
wherein R isTObeying a Bernoulli (Bernoulli) distribution function, p being the dropout regularization ratio; the neuron of each long-time memory neural network comprises a selection and conversion mechanism of three information of an input gate, a forgetting gate and an output gate, and the implementation process is as follows:
the forgetting gate determines whether the information from the previous neuron is retained, and takes a value of 1 or 0, wherein the former represents retention, and the latter represents discarding, as shown in formula (5):
Figure BDA0002548270080000064
wherein WfvAnd Wfh is a weight matrix of the input value of the current neuron and the output value of the previous neuron of the forgetting gate function, bfIs a bias parameter matrix;
input gateComputing retention information r for current neuron inputTAnd new state information sTAs shown in formulas (6) and (7):
Figure BDA0002548270080000065
Figure BDA0002548270080000066
wherein, WrvAnd WsvRespectively, a function r in the input gateTAnd sTWeight matrix of current neuron input values, WrhAnd WshRespectively, a function r in the input gateTAnd sTWeight matrix of previous neuron output values, brAnd bsRespectively, a function r in the input gateTAnd sTA bias parameter matrix of (a);
then the state c of long-time and short-time memory neural network neuronsTUpdating is done as equation (8):
cT＝rT⊙sT+cT-1⊙fT (8)
the output gate calculates the output information h of the current neuronTAs shown in formulas (9) and (10):
Figure BDA0002548270080000071
hT＝oT⊙tanh(cT) (10)
wherein WovAnd WohAs a function o in the output gateTWeight matrix of current neuron input values and previous neuron output values, boAs a function oTA bias parameter matrix of (a);
updating a weight matrix omega and a bias parameter matrix b of the long-time memory neural network by using an RMSprop algorithm according to the following formula (11-14):
Figure BDA0002548270080000072
Figure BDA0002548270080000073
Figure BDA0002548270080000074
Figure BDA0002548270080000075
wherein J1To calculate the cost function of the weight matrix, J2To calculate the cost function of the bias parameter matrix, k represents the number of current iterations, β is a coefficient, typically taken as 0.999, and ε is used to prevent the divisor from being zero, typically taken as 10-8Alpha is the learning rate, the algorithm belongs to a gradient descent acceleration method, so that the training of the memory neural network is ended after 20 iterations, and the model is stored; and inputting known data to obtain an open-circuit voltage prediction sequence.
S2: dividing the prediction result into a plurality of discharge cycles in sequence, and counting the number N of open-circuit voltage samples from the initial voltage to the minimum voltage in each discharge cycleSUsing the sampling time TSThe same, the time T to discharge to the minimum voltage in each discharge cycle is obtainedminThe method comprises the following specific steps:
s21: the method for sequentially dividing the prediction result into a plurality of discharge cycles is as follows:
if the time TjSo that equation (15) holds:
Figure BDA0002548270080000076
then Tj(j＝0，1，2，…，n；T01) is the end time of the jth cycle, and the open circuit voltage sequence of the jth cycle is
Figure BDA0002548270080000081
S22: counting the number N of open-circuit voltage samples from the initial voltage to the minimum voltage in each discharge cycleSUsing the sampling time TSThe same, the time T to discharge to the minimum voltage in each discharge cycle is obtainedminAs shown in formula (16):
Tmin＝NS×TS (16)
s3: establishing a capacity prediction model based on an artificial neural network to predict the capacity C of the lithium battery so as to obtain a predicted value RUL of the residual life of the lithium battery, and specifically comprising the following steps:
s31: setting the number of input neurons to NI2Setting the number of the nerve hiding layers to be LH2And number of hidden layer neurons NH2Establishing the time T from discharge to minimum voltage in the discharge cycleminMapping relation to lithium battery capacity C;
s32: the calculation process of each neuron of the hidden layer is shown as the formula (17):
HT＝σ(WHTTmin+bH)； (17)
HTrepresenting a hidden layer computation function, where WHTAnd bHInputting a weight matrix and a bias parameter matrix of the function input value; after 20 times of iteration, the training is finished, and the model is saved. The predicted time T for discharging to the minimum voltage in each discharge cycleminInputting a model to obtain a capacity prediction result;
s33: defining EOL as capacity CjEqual to rated capacity C0The number j of cycles corresponding to 70%; defining and calculating the residual life RUL of the lithium battery, as shown in formula (18):
RUL＝UL-EOL； (18)
wherein, UL is the number of cycles that the current lithium battery has been operated, and EOL is the end of life of the lithium battery.
The invention utilizes Python language to carry out modeling, respectively verifies the prediction results of the method for the lithium battery Life attenuation data provided by NASA Ames computers of Excellence (PCoE) and Center for Advanced Life Cycle Engineering (CALCE), and compares the prediction results with the prediction results of a long-time memory-artificial neural network coupling model and a simple long-time memory neural network model which adopt discharge time as characteristic parameters, each group of experiments adopt 30% of data for training, the prediction results are shown in figures 2 and 3 (the figure 2 is different from the lithium battery sample used in figure 3), wherein M1 is the prediction result of the long-time memory-artificial neural network coupling model which adopts open-circuit voltage as characteristic parameters, M2 is the prediction result of the long-time memory-artificial neural network coupling model which adopts discharge time as characteristic parameters, M3 is the prediction result of the simple long-time memory neural network model, the long-time and short-time memory-artificial neural network coupling model which takes the open-circuit voltage as the characteristic parameter has better prediction effect compared with other models, and the long-time and short-time memory-artificial neural network coupling model which takes the open-circuit voltage as the characteristic parameter has good generalization performance by predicting different types of batteries. To further illustrate, the prediction results are evaluated by using Absolute Error (AE) and Root Mean Square Error (RMSE), and the calculation process is shown in equations (19) and (20):
Figure BDA0002548270080000091
Figure BDA0002548270080000092
wherein RUL is the true value of the remaining cycle life of the battery,
Figure BDA0002548270080000093
as a prediction of the remaining cycle life of the battery, CjThe true value of the jth cycle battery capacity,
Figure BDA0002548270080000094
is a predicted value of the j-th cycle battery capacity. The results of the calculations are shown in fig. 4 and 5 (fig. 4 is the absolute error of the remaining life prediction, fig. 5 is the root mean square error of the capacity fade prediction),
the long-short time memory-artificial neural network coupling model established by the invention takes the open-circuit voltage as the characteristic parameter has small error of the prediction of the residual life of the lithium battery, has higher fitting precision of an actual capacity attenuation curve, and has very high residual life prediction precision.
The above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.

Claims (2)

1. A method for predicting the remaining life of a lithium battery of an electric forklift based on multi-neural network coupling is characterized by comprising the following steps:
s1: establishing an open-circuit voltage prediction model based on a long-time and short-time memory neural network, and optimizing the network by adopting an RMSprop algorithm and a dropout regularization method, thereby predicting the open-circuit voltage value V of the lithium battery in a discharge cycleOC；
S11: setting the number of input neurons to NI1Setting the number of the nerve hiding layers to be LH1And number of hidden layer neurons NH1Establishing a relation mapping between input parameters and output parameters by using a sliding window, and establishing an input V during training if the length of the sliding window is L and the length of training data is MinputTo outputs of V respectivelyoutputThe mapping relation of (2) is shown in formula (1):
Figure FDA0003194685840000011
wherein, VinputEach row in the set corresponds to an input sample with a length of L;
s12: model obtained by training S11 is utilized to split open-circuit voltage VOCThe prediction is carried out by adopting the following method, wherein the input and the output are shown as the formula (2):
Figure FDA0003194685840000012
wherein, VOCEach row in the circuit diagram corresponds to a predicted input sample, and each prediction obtains a predicted value of the open-circuit voltage
Figure FDA0003194685840000013
Using the obtained predicted values
Figure FDA0003194685840000014
Updating the input sample sequence for the next prediction, i.e. the predicted value
Figure FDA0003194685840000015
Adding the data to the end of the input sample sequence, and deleting the first data of the input sample sequence, thereby completing the window sliding operation and obtaining all predicted values of the open-circuit voltage
Figure FDA0003194685840000016
S2: dividing the prediction result into a plurality of discharge cycles in sequence, and counting the number N of open-circuit voltage samples from the initial voltage to the minimum voltage in each discharge cycleSUsing the sampling time TSThe same, the time T to discharge to the minimum voltage in each discharge cycle is obtainedmin；
S3: establishing a capacity prediction model based on an artificial neural network to predict the capacity C of the lithium battery so as to obtain a predicted value RUL of the residual life of the lithium battery;
s31: setting the number of input neurons to NI2Setting upThe number of the nerve hidden layers is LH2And number of hidden layer neurons NH2Establishing the time T from discharge to minimum voltage in the discharge cycleminMapping relation to lithium battery capacity C;
s32: the calculation process of each neuron of the hidden layer is shown as the formula (17):
HT＝σ(WHTTmin+bH)； (17)
HTrepresenting a hidden layer computation function, where WHTAnd bHInputting a weight matrix and a bias parameter matrix of the function input value; after 20 times of iteration, ending the training and storing the model; the predicted time T for discharging to the minimum voltage in each discharge cycleminInputting a model to obtain a capacity prediction result;
s33: defining EOL as capacity CjEqual to rated capacity C0The number j of cycles corresponding to 70%; defining and calculating the residual life RUL of the lithium battery, as shown in formula (18):
RUL＝UL-EOL； (18)
wherein, UL is the number of the operating cycles of the current lithium battery, and EOL is the end point of the service life of the lithium battery.
2. The method for predicting the residual life of the lithium battery of the electric forklift based on the multi-neural-network coupling as claimed in claim 1, wherein the specific steps of S2 are as follows:
s21: the method for sequentially dividing the prediction result into a plurality of discharge cycles is as follows:
if the time TjSo that equation (15) holds:
Figure FDA0003194685840000031
then Tj(j＝0，1，2，…，n；T01) is the end time of the jth cycle, and the open circuit voltage sequence of the jth cycle is
Figure FDA0003194685840000032
S22: counting the number N of open-circuit voltage samples from the initial voltage to the minimum voltage in each discharge cycleSUsing the sampling time TSThe same, the time T to discharge to the minimum voltage in each discharge cycle is obtainedminAs shown in formula (16):
Tmin＝NS×TS (16)。
CN202010568146.1A 2020-06-19 2020-06-19 Electric forklift lithium battery residual life prediction method based on multi-neural network coupling Active CN111948563B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010568146.1A CN111948563B (en) 2020-06-19 2020-06-19 Electric forklift lithium battery residual life prediction method based on multi-neural network coupling

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010568146.1A CN111948563B (en) 2020-06-19 2020-06-19 Electric forklift lithium battery residual life prediction method based on multi-neural network coupling

Publications (2)

Publication Number Publication Date
CN111948563A CN111948563A (en) 2020-11-17
CN111948563B true CN111948563B (en) 2021-12-17

Family

ID=73337098

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010568146.1A Active CN111948563B (en) 2020-06-19 2020-06-19 Electric forklift lithium battery residual life prediction method based on multi-neural network coupling

Country Status (1)

Country Link
CN (1) CN111948563B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112731184B (en) * 2020-12-28 2023-03-03 深圳供电局有限公司 Battery service life detection method and system
CN112986827B (en) * 2021-04-12 2022-06-03 山东凯格瑞森能源科技有限公司 Fuel cell residual life prediction method based on deep learning
CN113359048A (en) * 2021-04-28 2021-09-07 中国矿业大学 Indirect prediction method for remaining service life of lithium ion battery
CN113777496B (en) * 2021-09-06 2023-10-24 北京化工大学 Lithium ion battery residual life prediction method based on time convolution neural network
CN115186370A (en) * 2022-05-18 2022-10-14 广东海洋大学 Engineering forklift transfer learning system based on deep learning training model
CN114859249B (en) * 2022-07-06 2022-10-18 苏州清研精准汽车科技有限公司 Method and device for detecting battery pack capacity

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3520886B2 (en) * 1996-03-08 2004-04-19 サンケン電気株式会社 Rechargeable battery status determination method
JP6250298B2 (en) * 2013-04-15 2017-12-20 新電元工業株式会社 Secondary battery life prediction system and secondary battery characteristic evaluation device
CN104849671B (en) * 2015-05-22 2017-07-11 大连理工大学 A kind of battery capacity detecting system based on combination neural net
CN107167741A (en) * 2017-06-06 2017-09-15 浙江大学 A kind of lithium battery SOC observation procedures based on neutral net
CN109343505A (en) * 2018-09-19 2019-02-15 太原科技大学 Gear method for predicting residual useful life based on shot and long term memory network
CN109993270B (en) * 2019-03-27 2022-11-15 东北大学 Lithium ion battery residual life prediction method based on gray wolf group optimization LSTM network
CN110542866B (en) * 2019-10-12 2023-04-07 上海新微技术研发中心有限公司 Method for estimating residual electric quantity parameter of battery
CN110687460B (en) * 2019-10-12 2023-02-24 上海新微技术研发中心有限公司 Soc estimation method

Also Published As

Publication number Publication date
CN111948563A (en) 2020-11-17

Similar Documents

Publication Publication Date Title
CN111948563B (en) Electric forklift lithium battery residual life prediction method based on multi-neural network coupling
CN112241608B (en) Lithium battery life prediction method based on LSTM network and transfer learning
CN109993270B (en) Lithium ion battery residual life prediction method based on gray wolf group optimization LSTM network
WO2021138925A1 (en) Lithium battery capacity estimation method based on improved convolution-long short term memory neural network
CN112036084B (en) Similar product life migration screening method and system
CN111832220A (en) Lithium ion battery health state estimation method based on codec model
CN111999648A (en) Lithium battery residual life prediction method based on long-term and short-term memory network
CN110082682B (en) Lithium battery state of charge estimation method
CN111736084B (en) Valve-regulated lead-acid storage battery health state prediction method based on improved LSTM neural network
CN111308375B (en) LSTM-FFNN-based electric forklift lithium ion battery health state prediction method
CN114325450A (en) Lithium ion battery health state prediction method based on CNN-BilSTM-AT hybrid model
CN114861527A (en) Lithium battery life prediction method based on time series characteristics
CN112734002B (en) Service life prediction method based on data layer and model layer joint transfer learning
CN114035098A (en) Lithium battery health state prediction method integrating future working condition information and historical state information
CN115616414A (en) Method and device for predicting remaining life long-time sequence of energy storage lithium battery
Zhang et al. Evolving Elman neural networks based state-of-health estimation for satellite lithium-ion batteries
CN115308606A (en) Lithium ion battery health state estimation method based on proximity features
CN112257348B (en) Method for predicting long-term degradation trend of lithium battery
CN113361692B (en) Lithium battery remaining life combined prediction method
CN116047314B (en) Rechargeable battery health state prediction method
CN113791351A (en) Lithium battery life prediction method based on transfer learning and difference probability distribution
CN115047350B (en) Digital-analog linkage based lithium ion battery remaining service life prediction method
CN114757340A (en) Lithium battery health state prediction method and system based on neural network integration
CN115372827A (en) Battery SOC prediction method based on long-time and short-time memory network
CN114415034A (en) Lithium ion battery capacity prediction method based on internal mechanism analysis

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
CB03 Change of inventor or designer information
CB03 Change of inventor or designer information

Inventor after: Tong Shuiguang

Inventor after: Tong Zheming

Inventor after: Li Yuansong

Inventor after: Miao Jiazhi

Inventor before: Tong Zheming

Inventor before: Miao Jiazhi

Inventor before: Tong Shuiguang

Inventor before: Li Yuansong

TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20211122

Address after: 310000 Yuhang Tang Road, Xihu District, Hangzhou, Zhejiang 866

Applicant after: ZHEJIANG University

Applicant after: Hangcha Group Co., Ltd

Address before: 310058 Yuhang Tang Road, Xihu District, Hangzhou, Zhejiang 866

Applicant before: ZHEJIANG University

GR01 Patent grant
GR01 Patent grant