CN105678077A - Online prediction method of power performance of lithium ion battery for hybrid power vehicle - Google Patents
Online prediction method of power performance of lithium ion battery for hybrid power vehicle Download PDFInfo
- Publication number
- CN105678077A CN105678077A CN201610009520.8A CN201610009520A CN105678077A CN 105678077 A CN105678077 A CN 105678077A CN 201610009520 A CN201610009520 A CN 201610009520A CN 105678077 A CN105678077 A CN 105678077A
- Authority
- CN
- China
- Prior art keywords
- lithium ion
- ion battery
- performance
- model
- soc
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
Landscapes
- Secondary Cells (AREA)
Abstract
The invention relates to an online prediction method of the power performance of a lithium ion battery for a hybrid power vehicle. Parameters for representing the power performance of the lithium ion battery and main factors influencing the power performance of the lithium ion battery are selected; the main factors are adjusted to simulate the operating work conditions of the lithium ion battery in the hybrid power vehicle in the actual running process; the correlated relation among the main factors and the correlated relation between the main factors and the parameters are analyzed; an online prediction model of the power performance of the lithium ion battery is established based on the principle of a support vector machine; the model parameters of the model are optimized by utilizing a grid search method; a part of experimental data serves as a sample to train the model, and the final online prediction model of the power performance of the lithium ion battery is determined; the final model is verified; online prediction of the power performance of the lithium ion battery for the hybrid power vehicle is achieved by adopting the model. The model is high in accuracy, the training time is short, the response speed is high, quick estimation of the power performance of the battery can be achieved, and the safe and reliable running of a vehicle is guaranteed.
Description
Technical field
The present invention relates to the on-line prediction method of hybrid vehicle lithium ion battery power-performance.
Background technology
Along with socioeconomic development, living standards of the people are largely increased, but bring the surge of energy demand simultaneously, and environmental problem also highlights day by day. Electric automobile because its demand for energy is low, pollute little and be widely popularized, alleviate Pressure on Energy, improve the exhaust pollution problems of conventional fuel oil automobile.
As one of main electric automobile type, hybrid vehicle adopts gasoline and lithium ion battery collectively as the power source of automobile, provides the power demand in vehicle travel process. The real vehicle operating condition change of hybrid electric vehicle is very violent, and therefore battery peak power performance in short-term is directly connected to the safety and reliability of vehicle. The real-time estimate of battery peak power performance is particularly significant for integrated vehicle control tactics, thus becoming one of the major function and research emphasis of battery management system (batterymanagementsystem, BMS).
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of on-line prediction method of hybrid vehicle lithium ion battery power-performance, to improve the function of battery management system, it is ensured that reliability in vehicle travel process and safety.
For solving above-mentioned technical problem, the present invention adopts following technical proposals;
A kind of on-line prediction method of hybrid vehicle lithium ion battery power-performance, comprises the following steps:
S1, choosing to characterize the parameter of lithium ion battery power-performance, described parameter is 10 seconds peak powers, is the battery performance number that just keeps 10 seconds invariable power discharge conditions.
S2, choose the principal element affecting lithium ion battery power-performance, including: battery temperature, battery SOC (state-of-charge) and the internal resistance of cell.
S3: for simulation lithium ion battery real working condition in real vehicle running, the artificial battery temperature adjusted in lithium ion battery test process. 8 different temperature spots are such as set, respectively-20 DEG C ,-10 DEG C, 0 DEG C, 10 DEG C, 20 DEG C, 30 DEG C, 40 DEG C and 50 DEG C.By arranging Temperature of Warm Case, then battery is put into incubator sufficient standing (standing more than four hours), make battery temperature keep consistent with Temperature of Warm Case.
By battery charging and discharging being carried out regulating cell SOC, lithium ion battery is made to be operated under different battery SOCs, for instance battery SOC includes 10%SOC, 20%SOC, 30%SOC, 40%SOC, 50%SOC, 60%SOC, 70%SOC, 80%SOC and 90%SOC.
The real working condition in real vehicle running is simulated by adjusting battery temperature and battery SOC. And the internal resistance of cell is considered as uncontrollable, by recording in experimentation and recording. Under each temperature value, test obtains lithium ion battery and is in the internal resistance of cell under 9 different battery SOCs and 10 seconds peak powers.
Some groups of experimental datas are obtained, for instance 120 groups of experimental datas by invariable power method of testing. Often group experimental data includes 10 seconds peak powers of principal element (battery temperature, battery SOC and the internal resistance of cell) and correspondence.
Dependency relation between the principal element that S4, analysis obtain, between principal element and 10 seconds peak powers, concrete grammar is to use Pearson's correlation coefficient. The Pearson's correlation coefficient linear correlation degree for weighing between principal element, between principal element and 10 seconds peak powers, the absolute value of Pearson's correlation coefficient, closer to 1, illustrates that dependency is more strong; Otherwise, correlation coefficient is closer to 0, then illustrate that degree of association is more weak.
S5, set up lithium ion battery power-performance on-line prediction model based on support vector machine principle. Concrete grammar is, using the input as model of battery temperature, battery SOC and the internal resistance of cell, 10 seconds peak powers are as the output of model. Then the Radial basis kernel function of support vector machine is chosen:
σ is kernel functional parameter, is also core width; xiFor mode input vector, xjVector is supported for model is existing. And give kernel functional parameter and penalty coefficient initial value.
S6, utilizing trellis search method that the model parameter of lithium ion battery power-performance on-line prediction model is optimized, described model parameter includes kernel functional parameter σ and penalty coefficient c.
S7, lithium ion battery power-performance on-line prediction model is trained by experimental data as sample, the goodness of fit of assessment lithium ion battery power-performance on-line prediction model and the accuracy of training result, establish final lithium ion battery power-performance on-line prediction model.
Preferably, randomly select 80 groups of experimental datas (training sample) model is trained.
S8, final lithium ion battery power-performance on-line prediction model is verified, it was demonstrated that this model has high accuracy, high efficiency and the advantage such as workable.
Preferably, for 120 groups of experimental datas that experiment obtains, remove 80 groups of experimental datas as training sample outside, remaining 40 groups of experimental datas (sample data) are as forecast sample, and the power of battery performance online forecast model that the present invention is set up is verified.
S9, adopt the on-line prediction of final lithium ion battery power-performance on-line prediction model realization hybrid vehicle lithium ion battery power-performance.
Hybrid vehicle lithium ion battery power-performance Forecasting Methodology based on support vector machine and grid search of the present invention, has the advantages that
The battery short-time rating performance prediction model training time that the present invention sets up is short, fast response time, workable, 10 seconds peak powers of Accurate Prediction lithium ion battery in violent driving process can be changed at hybrid electric vehicle operating condition, realize the battery management system on-line prediction to power of battery performance, thus providing the control strategy of car load power distribution, it is ensured that the safe and reliable operation of hybrid electric vehicle.
Accompanying drawing explanation
The present invention has drawings described below:
Fig. 1 is the on-line prediction block diagram of lithium ion battery power-performance;
Fig. 2 is lithium ion battery invariable power test curve figure;
Fig. 3 is lithium ion battery invariable power test curve fitting result figure;
Fig. 4 is support vector machine nonlinear regression schematic diagram;
Fig. 5 is grid data service flow chart.
Detailed description of the invention
Below in conjunction with one group of embodiment and accompanying drawing, the present invention is described further.
The technical solution adopted for the present invention to solve the technical problems is a kind of on-line prediction method of hybrid vehicle lithium ion battery power-performance. The on-line prediction block diagram of hybrid vehicle lithium ion battery power-performance proposed by the invention as it is shown in figure 1, include adopts supporting vector machine model, mode input, model output, the training of model optimization method, model and checking. Using the lithium manganate battery of hybrid vehicle as object of study, set up the on-line prediction model of the type power of battery performance.
1, choose 10 seconds peak powers and characterize the power-performance of battery
Using lithium manganate battery as hybrid electric vehicle (hybridelectricvehicle, HEV) with lithium ion battery, according to USABC (UnitedStatesAdvancedBatteryConsortium) battery testing standard, its peak power persistent period is typically not greater than 10s. Voltage restricting condition for use according to battery, cell charging/discharging peak power is defined as terminal voltage after battery continues the charge/discharge t second under firm power and just arrives the highest/minimum running voltage that battery allows, then this constant power level is battery t second charge/discharge peak power, therefore chooses 10s peak power to characterize the power-performance of battery.
2, the principal element affecting lithium ion battery power-performance is chosen
In the actual travel process of hybrid electric vehicle, battery temperature, battery charge state and the internal resistance of cell directly affect the power-performance of battery, and and present obvious non-linear relation between 10 seconds peak powers of battery.
3, contrived experiment simulation lithium ion battery operating condition in hybrid electric vehicle actual travel process
For simulating the operating mode of lithium ion battery acute variation in hybrid electric vehicle actual travel process, the artificial ambient temperature adjusted in battery testing process, make battery operated under the duty of low temperature, room temperature and high temperature. Specifically have chosen 8 test temperature spots, be respectively as follows :-20 DEG C ,-10 DEG C, 0 DEG C, 10 DEG C, 20 DEG C, 30 DEG C, 40 DEG C and 50 DEG C. Simultaneously, make battery operated at different state-of-charge (stateofcharge by discharge and recharge adjustment, SOC) under condition, including 10%SOC, 20%SOC, 30%SOC, 40%SOC, 50%SOC, 60%SOC, 70%SOC, 80%SOC and 90%SOC. Such test condition arrange cover hybrid electric vehicle in actual motion faced by all real working condition so that experimental result has cogency. Experiment obtains 120 groups of experimental datas, and often group packet is containing four values, respectively 10 seconds peak powers of battery temperature, battery SOC, the internal resistance of cell and battery.
4, the dependency relation between analyzing influence factor and power of battery performance
Utilizing statistical method can study the dependency relation between a stochastic variable and multiple stochastic variable, set the variable of independent variable group as battery temperature, battery SOC and the internal resistance of cell, the variable of dependent variable group is 10 seconds peak powers of battery. Scatterplot is utilized to present the statistical relationship between independent variable group and dependent variable group intuitively.For embodying the dependency between variable more accurately, calculating the Pearson's correlation coefficient between any two variablees, result is as shown in table 1. Pearson's correlation coefficient is used to weigh the linear correlation degree between two group data sets, and the absolute value of Pearson's correlation coefficient, closer to 1, illustrates that dependency is more strong; Otherwise, correlation coefficient is closer to 0, then illustrate that degree of association is more weak.
Table 1 Pearson came correlation analysis
According to table 1, it is possible to analyze battery temperature, battery SOC, the internal resistance of cell and 10 seconds peak power degrees of correlation between any two and the significance between them. Significance refers to the probability being absent from linear relationship between two parameter, embodies the ability being mutually distinguishable between data. Obviously, significance is more little, illustrates between two parameter more relevant. If the correlation coefficient between battery temperature and the internal resistance of cell is-0.784, it is known that when confidence level (bilateral) is 0.01, both degrees of correlation are very big, again owing to the significance of statistical test is 0, less than 0.01, illustrate that dependency is very strong, and be negative correlation. Such as correlation coefficient between the internal resistance of cell and 10 seconds peak powers is-0.606 again, and the significance of statistical test, less than 0.01, illustrates that the two dependency is relatively big, is also negative correlation. In like manner can analyze the dependency relation between its dependent variable in 120 groups of data that experiment records.
5, the 10s peak power forecast model of battery is set up based on support vector machine principle
Known based on above-mentioned correlation analysis result, between variable and the variable of dependent variable group of independent variable group, all there is bigger dependency, therefore the input as model of battery temperature, battery SOC and the internal resistance of cell is chosen, 10 seconds peak powers of battery, as the output of model, set up model according to the ultimate principle of support vector machine.
Support vector machine is the machine learning algorithm of a kind of Corpus--based Method theory, the input space is transformed to a higher dimensional space by the nonlinear transformation that it is defined by interior Product function and kernel function, namely by kernel function, sample space is mapped to a high-dimensional feature space, this higher dimensional space is tried to achieve an optimal classification surface, and then obtains a kind of non-linear relation between input variable and output variable. Algorithm of support vector machine can regard a convex double optimization problem as, is converted by kernel function, it is achieved utilizes linear learning machine to solve the nonlinearity regression problem in sample space in feature space. The algorithm on support vector machine of choosing of kernel function has important impact. As a rule, there is a most suitable kernel function in each regression estimates system so that it is regression effect is best. Kernel function conventional at present has following several:
1) q order polynomial kernel function:
K(xi,xj)=[< xi,xj>+1]q(2)
In formula, q is degree of freedom, xiFor mode input vector, xjVector is supported for model is existing.
2) radial direction base (radialbasisfunction, RBF) kernel function
In formula, σ is kernel functional parameter, is also core width; xiFor mode input vector, xjVector is supported for model is existing.
3) Sigmoid kernel function
In formula, σ is kernel functional parameter, is also core width; xiFor mode input vector, xjVector is supported for model is existing.
The present invention is by contrasting three kinds of conventional kernel functions, it has been found that in regression estimates is applied, the performance of Radial basis kernel function is better than other kernel functions. Meanwhile, compared with other kernel functions, the form of Radial basis kernel function is simple, and parameter is few, and the fluctuation that predicts the outcome is little, and slickness is good.Therefore, this function kernel function as power of battery performance prediction supporting vector machine model is chosen.
6, trellis search method is adopted to optimize kernel functional parameter and penalty coefficient
After model elaborates, it is necessary to the parameter of model is optimized, the parameter object of optimization includes kernel functional parameter σ and penalty coefficient c. Kernel functional parameter σ generally adopts the mode of tentative calculation to obtain. Choosing σ in advance is a certain fixed value, then carries out model calculating, adjusts the value of σ according to result of calculation and experience, until obtaining the result of calculation the most satisfied that precision of prediction is the highest, it is believed that kernel functional parameter σ now is consistent with requirement. And penalty coefficient c controls the punishment degree beyond error sample, c value is more big then more strong to the binding character of sample, calculates process more complicated, calculates the time more long; Whereas if c value is too little, then causing that the accuracy of model and model accuracy can not be guaranteed, result loses contact with reality, and does not have cogency. Shortcoming that is low for algorithm above computational efficiency, that find global optimum difficulty, it is considered to the method introducing grid search (Grid-Search) obtains the global optimum of σ and c. Compare other parameter optimization methods two kinds conventional: genetic algorithm (Genetic-Algorithm) and chaotic optimization algorithm (Chaos-Optimization-Algorithm), grid-search algorithms is more suitable for small sample forecast model, although hunting zone is big, search speed is slower, but 2 parameter values can be simultaneously scanned for, it is ensured that obtain the optimized parameter that classification accuracy is the highest. Each group mutual decoupling of parameter in calculating process simultaneously, it is simple to parallel computation, operational efficiency is high.
7, according to experimental data model it is trained and verifies
The training of model
Randomly select 80 groups of experimental datas (training sample) model is trained, the model parameter (kernel functional parameter σ and penalty coefficient c) after being optimized based on grid data service respectively 4 and 16. Training result shows that model has significantly high precision, and the predictive value of model differs only small with the actual power value recorded. Data analysis is it can be seen that the predictive value of training sample is 41.69W with the mean error of the actual power value recorded by experiment, is the 2.61% of maximum peak power, and average relative error is 0.287; Maximum error 88.35W, is the 5.53% of maximum peak power. Owing to the peak power value of battery is general all relatively larger, the order of magnitude is in kW (kilowatt) left and right, therefore individual other extreme small (order of magnitude is at about ten watts) can widen mean error, causes the higher situation even greater than 100% of relative error. But it practice, the precision of whole model is significantly high, the predictive value of model and actual power value have the very strong goodness of fit, and variation tendency is extremely identical. After rejecting the data of low peak power (< 100W) again, relative error declines to a great extent, and meansigma methods is 2.61%. Meanwhile, this model has significantly high coefficient of determination (coefficientofdetermination, COD), and COD value, closer to 1, illustrates that this Regression Model Simulator data degree is more good, and result of calculation COD=0.9878, also demonstrate this point.
The checking of model
For experiment obtain 120 groups of experimental datas, remove 80 groups of experimental datas as training sample outside, remaining 40 groups of experimental datas (sample data) are as forecast sample. Sample data is predicted by the supporting vector machine model proposed by the present invention, and the peak power predictive value obtained and actual power value are also sufficiently close to.Peak power predictive value is fairly good with the degree of following of actual power value, and variation tendency also keeps consistent, and two curves degree are higher.
By the mean error of calculating, peak power predictive value and actual power value and maximum error respectively 59.16W and 189.53W, account for the 3.65% and 11.70% of maximum peak power. According to statistic analysis result, the estimation peak power of model and the relative error of true peak power number of batteries within 20% are 20, account for the 50% of sample data sum; 16, relative error battery within 10%, accounts for the 40% of sample data sum; 8, relative error battery within 5%, accounts for the 20% of sample data sum. Occurring that the reason of this situation is in that, the low SOC of low temperature causes the value of peak power only small, and at this moment slightly the percentage ratio of deviation relative error will be very big. The value that relative error is bigger is distributed in the part that peak power is low more. After the data rejecting low peak power (< 100W), now absolute error meansigma methods is 29.78W, and relative error is lower than 10%.
Model training process and the COD of proof procedure model, mean error and maximum error, as shown in table 2.
The model of the automobile-used lithium ion battery of hybrid electrically that the above-mentioned analytic explanation present invention proposes peak power prediction in 10 seconds has higher precision and accuracy rate, applicable ability for high-peak power occasion to be significantly larger than low peak power occasion, and this suitability being follow-up peak power Forecasting Methodology and comparative study provide to be instructed and reference.
Table 2SVM model application condition in training and proof procedure
8, the on-line prediction of this model realization hybrid vehicle lithium ion battery power-performance is adopted.
In sum, technical scheme of the present invention analyzes the non-linear relation between the multiple parameters affecting battery peak power in short-term in hybrid electric vehicle driving process according to Principle of Statistics, obtain the correlation coefficient between two between parameter, specify that the single influence factor influence degree to peak power, it is achieved that Decoupled; Set up battery peak power forecast model in short-term based on support vector machine principle, and with trellis search method, model parameter is optimized; Carry out the experiment under different temperatures and different state-of-charge combination for hybrid vehicle lithium manganate battery, simulate the real working condition in vehicle actual travel process; Experimental data demonstrates the accuracy of model.
Model prediction result show the performance number of model prediction and actual value to follow degree significantly high, mean error 3.65%, most high level error only 11.70%, and the bigger value of relative error is distributed in the part that peak power is low more, this is owing in low temperature low SOC situation, the peak power value in short-term of battery is only small, and slightly deviation will result in relative error surge; When peak power is higher than 100W, the relative error of model significantly reduces, and is maintained at less than 10%.
The present invention propose the method training time is short, fast response time, ensureing that low error, high-precision basis also have very strong operability, the Fast estimation to extensive battery peak power can be realized, provide powerful guarantee for hybrid electric vehicle safe and reliable operation.
Obviously; the above embodiment of the present invention is only for clearly demonstrating example of the present invention; and be not the restriction to embodiments of the present invention; for those of ordinary skill in the field; can also make other changes in different forms on the basis of the above description; here cannot all of embodiment be given exhaustive, every belong to apparent change or variation that technical scheme extended out still within protection scope of the present invention.
The content not being described in detail in this specification belongs to the known prior art of professional and technical personnel in the field.
Claims (7)
1. an on-line prediction method for hybrid vehicle lithium ion battery power-performance, comprises the following steps:
S1, choose to characterize the parameter of lithium ion battery power-performance; Described parameter is 10 seconds peak powers;
S2, choose the principal element affecting lithium ion battery power-performance; Described principal element includes: battery temperature, battery SOC and the internal resistance of cell;
S3, simulate lithium ion battery operating condition in hybrid electric vehicle actual travel process by adjusting the value of battery temperature and battery SOC; And adopt invariable power method of testing to obtain some groups of experimental datas, often group experimental data includes principal element and 10 seconds peak powers;
Dependency relation between the principal element that S4, analysis obtain, between principal element and 10 seconds peak powers;
S5, set up lithium ion battery power-performance on-line prediction model based on support vector machine principle;
S6, utilize trellis search method that the model parameter of lithium ion battery power-performance on-line prediction model is optimized; Described model parameter includes kernel functional parameter σ and penalty coefficient c;
S7, lithium ion battery power-performance on-line prediction model is trained by some experimental data as sample, the goodness of fit of assessment lithium ion battery power-performance on-line prediction model and the accuracy of training result, establish final lithium ion battery power-performance on-line prediction model;
S8, adopt the on-line prediction of final lithium ion battery power-performance on-line prediction model realization hybrid vehicle lithium ion battery power-performance.
2. the on-line prediction method of hybrid vehicle lithium ion battery power-performance as claimed in claim 1, it is characterised in that described battery temperature includes :-20 DEG C ,-10 DEG C, 0 DEG C, 10 DEG C, 20 DEG C, 30 DEG C, 40 DEG C and 50 DEG C.
3. the on-line prediction method of hybrid vehicle lithium ion battery power-performance as claimed in claim 1, it is characterized in that, described battery SOC includes: 10%SOC, 20%SOC, 30%SOC, 40%SOC, 50%SOC, 60%SOC, 70%SOC, 80%SOC and 90%SOC.
4. the on-line prediction method of hybrid vehicle lithium ion battery power-performance as described in any one of claim 1-3, it is characterized in that: under each temperature value, test obtains lithium ion battery and is in the internal resistance of cell under different battery SOC and characterizes the parameter of lithium ion battery power-performance.
5. the on-line prediction method of hybrid vehicle lithium ion battery power-performance as claimed in claim 1, it is characterised in that: by the dependency relation described in Pearson's correlation coefficient analytical procedure S4.
6. the on-line prediction method of hybrid vehicle lithium ion battery power-performance as claimed in claim 1, it is characterized in that: in step S5, the concrete grammar setting up lithium ion battery power-performance on-line prediction model is, using the principal element input as model, 10 seconds peak powers are as the output of model; Then choose the Radial basis kernel function of support vector machine, and give kernel functional parameter σ and penalty coefficient c initial value.
7. the on-line prediction method of hybrid vehicle lithium ion battery power-performance as claimed in claim 6, it is characterised in that: the Radial basis kernel function of described support vector machine is:
σ is kernel functional parameter, xiFor mode input vector, xjVector is supported for model is existing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610009520.8A CN105678077A (en) | 2016-01-07 | 2016-01-07 | Online prediction method of power performance of lithium ion battery for hybrid power vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610009520.8A CN105678077A (en) | 2016-01-07 | 2016-01-07 | Online prediction method of power performance of lithium ion battery for hybrid power vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105678077A true CN105678077A (en) | 2016-06-15 |
Family
ID=56299293
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610009520.8A Pending CN105678077A (en) | 2016-01-07 | 2016-01-07 | Online prediction method of power performance of lithium ion battery for hybrid power vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105678077A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106249159A (en) * | 2016-08-03 | 2016-12-21 | 北京品驰医疗设备有限公司 | The electric quantity monitoring method of a kind of brain pacemaker and electric quantity monitoring system |
| CN108287310A (en) * | 2017-01-09 | 2018-07-17 | 宝山钢铁股份有限公司 | A kind of lithium battery load detection device and Load Calculation Method |
| CN108549779A (en) * | 2018-04-20 | 2018-09-18 | 广汽丰田汽车有限公司 | Performance simulation computational methods, terminal device and the storage medium of hybrid vehicle |
| CN108681742A (en) * | 2018-04-12 | 2018-10-19 | 上海交通大学 | A kind of analysis method for analyzing driver driving behavioural characteristic sensibility |
| CN108944945A (en) * | 2018-07-10 | 2018-12-07 | 深圳地平线机器人科技有限公司 | Trend prediction method, device, electronic equipment and vehicle for assisting driving |
| CN111024103A (en) * | 2018-10-10 | 2020-04-17 | 现代自动车株式会社 | Route guidance device and method for electric vehicle |
| CN111157902A (en) * | 2020-01-06 | 2020-05-15 | 上海度普新能源科技有限公司 | Lithium ion battery peak power testing method and system |
| CN111999651A (en) * | 2020-07-20 | 2020-11-27 | 武汉数值仿真技术研究院有限公司 | Power performance test method for lithium ion battery for pure electric vehicle |
| CN113935225A (en) * | 2020-06-29 | 2022-01-14 | 中国科学院大连化学物理研究所 | Flow battery stack optimization and performance prediction method based on machine learning |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103487759A (en) * | 2013-09-18 | 2014-01-01 | 江苏大学 | Hybrid electric vehicle battery SOC prediction method |
| CN103487755A (en) * | 2012-06-08 | 2014-01-01 | 通用电气公司 | System and method for predicting power transmission performance of battery |
| CN104374998A (en) * | 2014-12-09 | 2015-02-25 | 安徽江淮汽车股份有限公司 | Power battery power test method and system |
-
2016
- 2016-01-07 CN CN201610009520.8A patent/CN105678077A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103487755A (en) * | 2012-06-08 | 2014-01-01 | 通用电气公司 | System and method for predicting power transmission performance of battery |
| CN103487759A (en) * | 2013-09-18 | 2014-01-01 | 江苏大学 | Hybrid electric vehicle battery SOC prediction method |
| CN104374998A (en) * | 2014-12-09 | 2015-02-25 | 安徽江淮汽车股份有限公司 | Power battery power test method and system |
Non-Patent Citations (2)
| Title |
|---|
| 孙丙香等: "基于ANFIS和减法聚类的动力电池放电峰值功率预测", 《电工技术学报》 * |
| 张向东等: "基于网格搜索的支持向量机砂土液化预测模型", 《应用力学学报》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106249159A (en) * | 2016-08-03 | 2016-12-21 | 北京品驰医疗设备有限公司 | The electric quantity monitoring method of a kind of brain pacemaker and electric quantity monitoring system |
| CN106249159B (en) * | 2016-08-03 | 2019-01-29 | 北京品驰医疗设备有限公司 | A kind of electric quantity monitoring method and electric quantity monitoring system of brain pacemaker |
| CN108287310A (en) * | 2017-01-09 | 2018-07-17 | 宝山钢铁股份有限公司 | A kind of lithium battery load detection device and Load Calculation Method |
| CN108287310B (en) * | 2017-01-09 | 2020-03-31 | 宝山钢铁股份有限公司 | Lithium battery load detection device and load calculation method |
| CN108681742A (en) * | 2018-04-12 | 2018-10-19 | 上海交通大学 | A kind of analysis method for analyzing driver driving behavioural characteristic sensibility |
| CN108681742B (en) * | 2018-04-12 | 2021-09-28 | 上海交通大学 | Analysis method for analyzing sensitivity of driver driving behavior to vehicle energy consumption |
| CN108549779B (en) * | 2018-04-20 | 2021-09-28 | 广汽丰田汽车有限公司 | Performance simulation calculation method of hybrid electric vehicle, terminal device and storage medium |
| CN108549779A (en) * | 2018-04-20 | 2018-09-18 | 广汽丰田汽车有限公司 | Performance simulation computational methods, terminal device and the storage medium of hybrid vehicle |
| CN108944945A (en) * | 2018-07-10 | 2018-12-07 | 深圳地平线机器人科技有限公司 | Trend prediction method, device, electronic equipment and vehicle for assisting driving |
| CN108944945B (en) * | 2018-07-10 | 2020-03-20 | 深圳地平线机器人科技有限公司 | State prediction method and device for driving assistance, electronic equipment and vehicle |
| CN111024103A (en) * | 2018-10-10 | 2020-04-17 | 现代自动车株式会社 | Route guidance device and method for electric vehicle |
| CN111157902A (en) * | 2020-01-06 | 2020-05-15 | 上海度普新能源科技有限公司 | Lithium ion battery peak power testing method and system |
| CN111157902B (en) * | 2020-01-06 | 2022-05-27 | 上海度普新能源科技有限公司 | Lithium ion battery peak power testing method and system |
| CN113935225A (en) * | 2020-06-29 | 2022-01-14 | 中国科学院大连化学物理研究所 | Flow battery stack optimization and performance prediction method based on machine learning |
| CN111999651A (en) * | 2020-07-20 | 2020-11-27 | 武汉数值仿真技术研究院有限公司 | Power performance test method for lithium ion battery for pure electric vehicle |
| CN111999651B (en) * | 2020-07-20 | 2023-06-06 | 武汉数值仿真技术研究院有限公司 | Power performance test method of lithium ion battery for pure electric automobile |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105678077A (en) | Online prediction method of power performance of lithium ion battery for hybrid power vehicle | |
| Lu et al. | Online estimation of state of power for lithium-ion batteries in electric vehicles using genetic algorithm | |
| Liu et al. | An advanced Lithium-ion battery optimal charging strategy based on a coupled thermoelectric model | |
| Shen et al. | Design and real-time controller implementation for a battery-ultracapacitor hybrid energy storage system | |
| Neubauer et al. | A second life for electric vehicle batteries: answering questions on battery degradation and value | |
| Onori et al. | A new life estimation method for lithium-ion batteries in plug-in hybrid electric vehicles applications | |
| Shen et al. | Fuzzy model based control for energy management and optimization in fuel cell vehicles | |
| Hamida et al. | Parameter identification and state of charge estimation of Li-Ion batteries used in electric vehicles using artificial hummingbird optimizer | |
| Eddahech et al. | Adaptive voltage estimation for EV Li-ion cell based on artificial neural networks state-of-charge meter | |
| Shi et al. | Electric vehicle battery remaining charging time estimation considering charging accuracy and charging profile prediction | |
| CN103983920A (en) | Method for establishing model of power battery of electric vehicle | |
| CN108804764A (en) | A kind of aging of lithium battery trend forecasting method based on extreme learning machine | |
| Liang et al. | Evaluation of battery modules state for electric vehicle using artificial neural network and experimental validation | |
| Shen et al. | Nonlinear optimization strategy for state of power estimation of lithium-ion batteries: A systematical uncertainty analysis of key impact parameters | |
| US11609273B2 (en) | Method and system for optimizing BMS model, storage medium and electric vehicle | |
| Kumari et al. | Electric vehicle battery state-of-charge estimation based on optimized deep learning strategy with varying temperature at different C Rate | |
| Zhang et al. | Optimization of Tremblay's battery model parameters for plug-in hybrid electric vehicle applications | |
| Gao et al. | Cyber Hierarchy Multiscale Integrated Energy Management of Intelligent Hybrid Electric Vehicles | |
| Li et al. | Joint estimation of SOC and SOH for lithium-ion batteries based on EKF multiple time scales | |
| Tong et al. | Comprehensive battery equivalent circuit based model for battery management application | |
| Upashrutti et al. | Estimation of State of Charge of EV Batteries-A Machine Learning Approach | |
| Karunarathna et al. | The Impact of Operational and Environmental Conditions on Battery Lifetime in Fast Electric Vehicle Charging Systems | |
| Liu et al. | Fault Diagnosis of Series Batteries based on GWO-SVM | |
| Karami et al. | ANN modeling of cold cranking test for sealed lead-acid batteries | |
| Hametner et al. | Model based lithium ion cell ageing data analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160615 |
|
| RJ01 | Rejection of invention patent application after publication |