DEVELOPMENT OF AN AUTOMATIC MEASUREMENT AND MATCHING MACHINE FOR COLUMNED BATTERY CELLS

Transcription

1 DEVELOPMENT OF AN AUTOMATIC MEASUREMENT AND MATCHING MACHINE FOR COLUMNED BATTERY CELLS Xiaoxing Li 1, Yi-Hua Fan 2, Ching-En Chen 2, Chia-Hui Tsao 2, Yu-Ming Chen 2 and Sheng-Chung Hsieh 1 1 School of Auto and Mechanical Engineering, Zhejiang Industry and Trade Vocational College, Wenzhou City, China 2 Department of Mechanical Engineering, Chung Yuan Christian University, Taiwan, R.O.C. yihuafan@cycu.edu.tw IMETI 2015 H5004_SCI No. 16-CSME-36, E.I.C. Accession 3922 ABSTRACT An automatic measurement and matching machine for columned batteries is developed in this study. The automatic machine consists of a battery feeding case, an intermittent separation feeding mechanism, a test region, eight classification transport channels, and battery depositary boxes. Besides the mechanism design, a human-machine interface and control program are written in VB language. The program provides the control and monitoring program for the auto-measurement system. The program cannot only read the measurement data and control the automatic machine, but also store these data and provide the test data histories for each cell supplier to the user. The experimental results show that the automatic machine could examine and classify the columned battery cells efficiently and decrease the demand of manpower. The results also show that the total measured and sorting quantity of one machine for eight hours is about 5760 pieces, which is greater than the 3500 pieces measured by one manpower. Keywords: columned batteries; automatic measurement and matching machine. DÉVELOPPEMENT D UN SYSTÈME AUTOMATIQUE DE MESURE ET D ÉQUIVALENCE POUR DES CELLULES DE BATTERIE EN COLONNE RÉSUMÉ Un système automatique de mesure et d équivalence pour les batteries en colonne est développé dans la présente étude. La machine automatique consiste d un bloc d alimentation de batterie, d un mécanisme d alimentation intermittent séparé, d une zone de test ; d une classification à huit voies d entrée et de caissons porte-batteries. Outre la conception du mécanisme, une interface homme-machine et un programme de commande sont écrits en langage VB. Le programme assure la commande et la surveillance du système automatique de mesure. Le programme peut non seulement lire les données de mesure et contrôler la machine, mais aussi il peut aussi conserver ces données et produire l historique des données à l utilisateur pour tous les fournisseurs de batteries. Les résultats expérimentaux démontrent que la machine automatique pouvait efficacement examiner et classifier les cellules des batteries en colonne et diminuer la demande en main d œuvre. Les résultats indiquent que la quantité totale de mesure et de triage d une machine pour huit heures est autour de 5760 pièces, ce qui est plus que les 3500 pièces mesurés par la main d œuvre. Mots-clés : batterie en colonne; machine automatique de mesure et de vérification. Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5,

2 1. INTRODUCTION A reduction in air pollution would result if more vehicles were to use gasoline/electric hybrid engines or run on pure electricity. The development of large-capacity batteries is important if electricity is to be efficiently stored. A high-capacity battery formed by one cell is difficult to produce, so most of the large-capacity batteries currently on the market are assembled using many small-capacity batteries in series or in parallel. Because the internal workings involve many small batteries connected in series or in parallel, in such largecapacity batteries the performance varies according to the performance of each of the small batteries. If one of the small batteries has performance issues, for example, having a larger internal resistance than the others, then this battery will suffer damage more readily and reduce the overall battery life. To avoid the problem of these small batteries negatively affecting the performance of the high-capacity battery, it is necessary to ensure that consistent performance of every small battery used in the large-capacity battery be maintained. In order to unify the performance of each batch of batteries, it is necessary to monitor the production of the small batteries. Battery cells with the same performance level must be selected in accordance with the measurement data of the voltage and internal resistance value in order to obtain a large-capacity battery comprising a group of batteries matched to ensure its performance. In this paper, an automated battery cell detection machine is proposed, the aim being to improve the detection efficiency for batteries, which is the current specification for the production of large-capacity batteries. The battery is the cylindrical type, which is the type most used to assemble the power modules of electric vehicle or notebook batteries. Many studies [1 5] have investigated methods of battery detection related to the use of computers to store the voltage, current, temperature and impedance value of the battery, as well as analyzing and determining battery life by these values. However, most of these studies focus on the establishment of inspection architectures; there are no studies which combine this approach with automatic detection of the battery system. The studies on automated testing systems have proposed different applications, such as the automatic assembly and measurement for the optical add/drop module of optical fibers and filters [6]. Quan et al. [7] used an area sensor-based robotic 3D measurement system, whereby the automotive part s surface is acquired patch-by-patch, to significantly reduce the inspection time. A 3D measurement system which uses an ATOS optical measurement instrument in a variety of experiments and the observation of objects has also been proposed in an attempt to develop a 3D optical measurement automated positioning system [8]. The automated battery testing machine proposed in the study used a feed mechanism to feed the battery cells to the detection region in order to measure the properties of the cells, and then separate them into several batches for further assembly work. The human-machine interface and control program are written in VB language. The program provides the monitoring and control functions for the auto-measurement system, and then through the feeding mechanism, measures the properties of batteries and classifies them to the classified boxes, thereby increasing the detection efficiency and reliability, as well as increasing the lifetime of products using in a large-capacity battery. The experimental results showed that the system could measure and sort 5760 pieces of battery cells in eight working hours. 2. MATERIAL AND METHODS 2.1. System Framework Figure 1 shows the framework of the proposed automatic measurement system. Our system design included control software running on a personal computer server that controls a battery tester through the serial port and all the mechanisms through a digital I/O card. The digital I/O card was used to either input or output digital signals from the surrounding mechanisms. Figure 2 shows the photograph of the automatic measurement and selection system for columned batteries. The system consisted of a battery feeding case, a separation feeding mechanism, a test region and a classification part with eight transport channels and battery depositary boxes. 694 Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5, 2016

3 Fig. 1. Framework of the automatic measurement system. Fig. 2. The automatic measurement and selection system for columned batteries. The battery feeding case is designed to place the unclassified battery cells. One case can place 120 pieces of battery cells maximum. As the case put on the top of the separate feeding mechanism, these battery cells will be fed to the detection region by the separate feeding mechanism. Figure 3 shows the structure and photograph of the separate feeding mechanism. It comprises a Geneva mechanism, a pair of gears, and a sprocket wheel. Considering the battery size and the desired categories, the sprocket wheel is divided into twenty sections. There is one notch on each section. The degrees between two notches are 18. As the wheel rotating, the battery cells will full into the top three notches from the feeding case. A five-stop Geneva mechanism, as shown in Fig. 3a, driven by a motor and combined a pair of gears with the teeth of the pinion and gear are 20 and 80 individually are proposed to form an intermittent motion device. The details of the mechanism are shown in Figs. 3b and 3c. The speed of the motor is set at 12 rpm, thus the intermittent motion device can drive the sprocket wheel to rotate 18 as the motor rotating one turn and feed the battery cells one by one to the test region to measure Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5,

4 (a) Intermittent motion device (b) Front view of the separate feeding mechanism (c) Back view of the separate feeding mechanism Fig. 3. Structure and photographs of the separate feeding mechanism. 696 Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5, 2016

5 the voltage and resistance in the speed of 5 sec/piece. The detection system uses an internal resistance meter to measure the voltage and resistance of the cell and transfer it to the computer to be classified by the program; the detection intervals were set as 3 mv and 3 mω. Then, the cell is transferred to the classification system by the sprocket wheel. According to measured data, the cell will be delivered to different transport channels by the solenoid with the computer command Automatic Measurement Software In this research, the human-machine interface and control program is written in VB language. The VB provided us with a simple and straight forward development environment capable of connecting the battery tester and all the mechanisms designed to meet the requirements of automatic examination for columned batteries of the type. The motion sequence diagram of the automation system is shown in Fig. 4. First is to place the batteries into the feeding case and put it on the separation feeding machine. Second is to decide the demand measured number of battery and desired categories. Turn on the motor to feed the battery cells one by one, and then to measure the properties of each battery cell and to sort it to the correct depositary box. The functions of the monitoring and control system were detecting, identifying, showing and saving the test signals of battery cells by the computer and to control the mechanism to sort the cells by the separation criterion which is set by the program or the operator. Besides controlling the auto-measurement machine, the man-machine interface program also provided a report system to show the test results in order to provide the operator with data on each battery cell, the properties, and the amount of each depositary box. There were three different operating programs for different purposes. The first one was designed to separate the new and untested batteries into eight categories. The criterion of the first category depends on the voltage (V 1) and internal resistance (R1) of the first tested battery, and set as {(V,R) V V 1 ± 15 mv,r R1 ± 1.5 mω}. As the first one battery for which the voltage and internal resistance values (V 2,R2) are not included in the region of the first criterion, they will be set as the second criterion of the second category and can be expressed as {(V,R) V V 2 ± 15 mv,r R2 ± 1.5 mω}. By the same method, as the first one battery for which the voltage and internal resistance values (V 3,R3) are not including in the region of the first and second criterions, they will be set as the third criterion of the third category and can be expressed as {(V,R) V V 3 ± 15 mv,r R3 ± 1.5 mω}. The criterions of categories four to seven can be chosen in the same manner. The eighth category included the batteries for which the measurement voltage and internal resistance do not including in the region of the first to the seventh categories. The second operating program was designed to subdivide the batteries in each category to six criterions after the cells have been classified by the first program. Thus the batteries in one of the first to the seventh categories can be subdivided to more accurate region. For example, if we want to subdivide the first category, the new criterions of the first six categories will be set as {(V,R) V [V 1,V mV],R [R1 1.5 mω,r1 0.5 mω)}, {(V,R) V [V 1,V mv],r [R1 0.5 mω,r mω)}, {(V,R) V [V 1,V mv],r [R mω,r mω)}, Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5,

6 Fig. 4. Framework of the automatic measurement system. 698 Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5, 2016

7 (a)human-machine interface. (b) Test report for all cells. Fig. 5. Human-machine interface and test reports. Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5,

8 (c) Batch report. Fig. 5. (Continued) {(V,R) V [V 1 15 mv,v 1),R [R1 1.5 mω,r1 0.5 mω)}, {(V,R) V [V 1 15 mv,v 1),R [R1 0.5 mω,r mω)}, {(V,R) V [V 1 15 mv,v 1),R [R mω,r mω)}. and The third operating program was designed to select the battery cells with special criterion and defined as the first criterion of the first category. The others are set as the criterion of the eighth category. The desired selection range can be set by the operator and the suitable batteries are feeding to the first depositary box and the other batteries are sent to the eighth depositary box. 3. RESULTS Figure 5a shows the human-machine interface of the case of the first operating program; Figs. 5b and c are the test reports for all the battery cells and the individual day reports. The test results show that the battery properties can be measured and classified to the right interval and the process time for one battery is about 5 seconds. Under such an examination speed, the total measured quantity of one machine in eight hours is about 5760 pieces, which is greater than the quantity of 3500 pieces measured by one manpower in the same work time. From the test result, it is obvious that the automatic measurement and matching machine could examine and sort the columned battery cells automatically and could also enhance the working efficiency and accuracy and decrease the demand of manpower. Furthermore, the automation system can save the test data and send out test reports if necessary. It is useful for the engineer or manager to check and confirm the property of the battery cells. CONCLUSIONS This study proposed a battery measurement and classification system for the columned battery cell of the type. The experimental results showed that the batteries were all classified correctly in the interval; the average time for one battery was about 5 seconds by the automatic measurement and matching machine. The total measured quantity for eight hours was about 5760 pieces, greater than the 3500 pieces measured by manpower. 700 Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5, 2016

9 ACKNOWLEDGEMENT This research was supported in part by the specific research fields project of the Chung Yuan Christian University, Taiwan, under grant NSC E CC3. REFERENCES 1. Singh, P., Fennie, C. Jr. and Reisner, D.E., A fuzzy system methodology to determine battery state-of-charge in primary Li/SO 2 and other batteries, in Proceedings of the 38th Power Sources Conference, Cherry Hill, NJ, pp. 295, June 8 11, Naoki, K. and Katsuhiko, Y., Estimation of the capacity of nickel-cadmium batteries by measuring impedance using electrolyte-deficient battery characteristics, in Proceedings of the 17th International Telecommunications Energy Conference, Hague, pp , October 29 November 1, Liu, Y.C., Battery management systems for improving battery efficiency in electric vehicles, World Electric Vehicle Journal, Vol. 4, pp , Chen, L., Zhang, Q., Fan, Y.Q., Wan, G.C. and Yin, G.H., A strategy of estimating stage-of-charge of LiFePO4 battery based on multirate extended kalman filter, in Proceedings of the 9th IEEE Conference on Industrial Electronics and Applications, Hangzhou, pp , June 9 11, Guo, X.W. Kang, L.Y. and Huang, Z.Z., On-line state of charge estimation of lithium-ion power battery pack using optimized unscented Kalman filtering, in Proceedings of the 2014 IEEE Transportation Electrification Asia-Pacific Conference and Expo, Beijing, pp. 1 6, August 31 September 3, Seo, Y.H. Park, H.S. Whang, K.H. Choi, D.S. and Je, T.J., Development of automatic assembly and evaluation systems for the optical add/drop module, in Proceedings of the 2004 ASME International Mechanical Engineering Congress and Exposition, Anaheim, pp , November 13 19, Shi, Q., Xi, N. and Chen, Y.F., Development of an automatic optical measurement system for automotive part surface inspection, in Proceedings of the 2005 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, Monterey, California, USA, Vol. 2, pp , GOM, ATOS Optical Measuring Machine User Manual, GOM Optical Measuring Techniques Inc., Transactions of the Canadian Society for Mechanical Engineering, Vol. 40, No. 5,