Integration of growing self-organizing map and continuous genetic algorithm for grading lithium-ion battery cells

doi:10.1016/j.asoc.2012.01.018

Applied Soft Computing

Volume 12, Issue 8, August 2012, Pages 2012-2022

https://doi.org/10.1016/j.asoc.2012.01.018 Get rights and content

Abstract

This study attempts to employ growing self-organizing map (GSOM) algorithm and continuous genetic algorithm (CGA)-based SOM (CGASOM) to improve the performance of SOM neural network (SOMnn). The proposed GSOM + CGASOM approach for SOMnn is consisted of two stages. The first stage determines the SOMnn topology using GSOM algorithm while the weights are fine-tuned by using CGASOM algorithm in the second stage. The proposed CGASOM algorithm is compared with other two clustering algorithms using four benchmark data sets, Iris, Wine, Vowel, and Glass. The simulation results indicate that CGASOM algorithm is able to find the better solution. Additionally, the proposed approach has been also employed to grade Lithium-ion cells and characterize the quality inspection rules. The results can assist the battery manufacturers to improve the quality and decrease the costs of battery design and manufacturing.

Graphical abstract

Highlights

► This study attempts to employ growing self-organizing map (GSOM) algorithm and continuous genetic algorithm (CGA)-based SOM (GASOM) algorithm to improve the performance of SOM neural network (SOMnn). ► The simulation results indicate that GASOM algorithm is able to find the better solution. ► The proposed approach has been also employed to grade lithium-ion cells and characterize the quality inspection rules. ► The results can assist the battery manufacturers to improve the quality and decrease the costs of battery design and manufacturing.

Introduction

In recent years, cluster analysis has been widely applied in many applications. Cluster analysis is the process of identifying natural groupings or clusters within multidimensional data, based on some similar measures, like Euclidean distance [1]. Its main purpose is to group samples with the same statistical characteristics together into the same cluster in order to achieve higher similarities within same clusters. Also, there are more significant differences between different clusters [2]. Cluster analysis is implemented by using statistical or neural network (NN) methods. Due to its high analytical value, cluster analysis is widely applied in variety of areas including business, education, social sciences, genetics and biology.

Among the major clustering algorithms, unsupervised neural network is one of the most representative methods while self-organizing map neural network (SOMnn) [3] is the most frequently applied one since it can be used for applications like image processing and mode identification. SOMnn mainly searches for different clusters within the data through sample training. Some researches have been made to propose different versions of SOMnns. Thus, this study aims to develop an enhanced cluster analysis approach which is able to group data which share similar characteristics to discover potential data traits and usable information based on SOMnn.

The proposed approach is a two-stage cluster analysis algorithm. In the first stage, the growing SOM (GSOM) algorithm [4] uses input data to determine the suitable SOMnn topology. In the second stage, continuous genetic algorithm (CGA) [5] is integrated with conventional SOM (i.e., CGA-based SOM (CGASOM)) algorithm in searching for the optimal weight vectors of SOMnn. To verify the proposed CGASOM approach for SOMnn, GSOM algorithm and four benchmark data sets (i.e., Iris, Glass, Vowel, and Wine) are employed. Then, the proposed CGASOM algorithm is further applied for grading batteries while incorporating real battery measurement results to obtain a comprehensive assessment. The result can be used as the foundation for automated battery grading evaluation system.

The rest of this paper is organized as follows. Section 2 presents the general background related to this study, while the proposed CGASOM algorithm is explained in Section 3. Sections 4 Simulation results and analysis, 5 Model evaluation results show the simulation results and the model evaluation results, respectively. The concluding remarks are finally made in Section 6.

Section snippets

Literature review

This section will briefly presents general backgrounds regarding self-organizing map (SOM) neural network, genetic algorithm and hybrid network.

Methodology

SOMnn has been widely applied in many areas recently since the algorithm is easily implemented. Additionally, SOMnn can visualize data clustering and obtain satisfactory results. However, for traditional SOMnn, the training is conducted by user-defined network topology. This may cause the need to try different network topologies for training. As a result, a lot of time will be spent for testing parameter settings. This study hence aims to improve SOMnn using a two-stage approach, GSOM + GASOM.

Simulation results and analysis

This study uses the GSOM algorithm first to determine SOMnn topology. Then, the attempt is made to compare three algorithms including CGA, CGASOM, and SOM + CGA. Thus, for these three algorithms, the network topology is from GSOM. The conceptual framework for these three algorithms is illustrated in Fig. 7. For GA, the initial population is randomly generated. For CGASOM, the initial population is also randomly generated. Then, in each iteration, CGA and SOM are implemented once. However, in SOM +

Model evaluation results

As portable electronic products get “lighter, thinner, shorter and smaller”, nowadays electronic components are also getting more and more compact. Particularly, small-sized, light weight high efficiency rechargeable batteries with high energy density have the highest demands. Assessing battery's performance has always been a major concern for battery manufacturers. To manufacture products with high efficiency usually means an increase in manufacturing cost. Therefore, it is essential to

Conclusions

This study first uses the GSOM algorithm to determine the SOMnn topology, and then compares CGA, CGASOM, and SOM + CGA algorithms utilizing four benchmark data sets. The computational results indicate that the proposed clustering algorithm yields the best performance, or lowest MAD value. This reveals that GSOM algorithm can take advantage of trait of CGA algorithm with the ability to escape from local optimal solution, and the self-clustering ability of SOMnn to carry out network training,

Acknowledgements

We would like to express our thanks to case company and Dr. S. C. Chi for providing the lithium-ion polymer battery data.

References (52)

R. Chelouah et al.
Genetic and Nelder–Mead algorithms hybridized for a more accurate global optimization of continuous multiminima functions
European Journal of Operational Research
(2003)
S.H. Liao et al.
Artificial neural networks classification and clustering of methodologies and applications—literature analysis from 1995 to 2005
Expert Systems with Applications
(2007)
T. Mu et al.
Breast cancer detection from FNA using SVM with different parameter tuning systems and SOM-RBF classifier
Journal of the Franklin Institute
(2007)
A. Forti et al.
Growing hierarchical tree SOM: an unsupervised neural network with dynamic topology
Neural Networks
(2006)
P.C. Chang et al.
Combining SOM and fuzzy rule base for flow time prediction in semiconductor manufacturing factory
Applied Soft Computing
(2006)
E.V. Samsonova et al.
TreeSOM: cluster analysis in the self-organizing maps
Neural Network
(2006)
B. Fritzke
Growing cell structures—a self-organizing network for unsupervised and supervised learning
Neural Networks
(1994)
S. Marsland et al.
A self-organizing network that grows when required
Neural Network
(2002)
J. Ontrup et al.
Large-scale data exploration with the hierarchically growing hyperbolic SOM
Neural Networks
(2006)
C.S. Ong et al.
Model identification of ARIMA family using genetic algorithm
Applied Mathematics and Computation
(2005)

S. Olafsson et al.

Operations research and data mining

European Journal of Operational Research

(2008)

R.J. Kuo et al.

Integration of ART2 neural network and genetic K-means algorithm for analyzing Web browsing paths in electronic commerce

Decision Support Systems

(2005)

R.J. Kuo et al.

Integration of self-organizing feature maps neural network and genetic K-means algorithm for market segmentation

Expert Systems with Applications

(2006)

K.J. Kim et al.

A recommender system using GA K-means clustering in an online shopping market

Expert Systems with Applications

(2008)

W. Yan et al.

An integration of bidding-oriented product conceptualization and supply chain formation

Computers in Industry

(2008)

J.J. Fuertes et al.

Visual dynamic model based on self-organizing maps for supervision and fault detection in industrial processes

Engineering Applications of Artificial Intelligence

(2010)

S. Delgado et al.

A combined measure for quantifying and qualifying the topology preservation of growing self-organizing maps

Neurocomputing

(2011)

P. Doganis et al.

Time series sales forecasting for short shelf-life food products based on artificial neural networks and evolutionary computing

Journal of Food Engineering

(2006)

A. Azadeh et al.

Integration of genetic algorithm, computer simulation and design of experiments for forecasting electrical energy consumption

Energy Policy

(2007)

R. Kubota et al.

Genetic algorithm with modified reproduction strategy based on self-organizing map and usable schema

International Congress Series

(2006)

R.J. Kuo et al.

Application of ant K-means on cluster analysis

Computers & Mathematics with Applications

(2005)

A.K. Jain et al.

Data clustering: a review

ACM Computer Survey

(1999)

J. Han et al.

Data Mining: Concepts and Techniques

(2006)

T. Kohonen

The self-organizing map

Proceedings of the IEEE

(1990)

L.D. Alahakoon et al.

Dynamic self-organizing maps with controlled growth for knowledge discovery

IEEE Transactions on Neural Networks

(2000)

C.C.H. Chan

Online auction customer segmentation using a neural network model

International Journal of Applied Science and Engineering

(2005)

Cited by (13)

A scalable multi-signal approach for the parallelization of self-organizing neural networks
2020, Neural Networks
Citation Excerpt :
One of the main fields of application of these networks is data clustering, in particular with big data that are difficult to manage with more traditional database tools and data processing applications. Over the years, SONNs have also been applied to more specific tasks such as image compression and/or segmentation(Kanimozhi & Bindu, 2013; Kulkarni, Forster, & Venayagamoorthy, 2011; Paul & Gupta, 2013), gesture recognition (Caridakis, Karpouzis, Drosopoulos, & Kollias, 2010), stereo-vision systems for robotics applications (Croom, Rucker, Romano, & Webster, 2010; Sharma, Kim, & Singh, 2012), wireless communication and sensor networks (Banković et al., 2010; Patra, Chattopadhyay, Bhaumik, & Roy, 2011), quality and cost optimization in design and manufacturing of goods (Kuo, Wang, & Chen, 2012) and many others. Among these applications, one common trait may be identified, that is the need for high performance computing, in particular in the presence of real-time requirements.
Self-Organizing Neural Networks (SONNs) have a wide range of applications with massive computational requirements that often need to be satisfied with optimized parallel algorithms and implementations. In literature, SONN have been generally parallelized with GPU computing according to a single-signal paradigm: each GPU thread manages one or more nodes of the network and works concurrently on one input signal at the time. This paper presents two contributions. The first one is the experimental proof that the single-signal approach for SONNs is not optimal for the task, as it is intrinsically sequential at its core and thus inherently limited in its performance. The non-optimality of the single-signal paradigm is illustrated via a specific and simplified benchmark. The second contribution is the introduction of a new multi-signal paradigm for the parallelization of SONNs, whereby multiple signals are processed at once in each iteration hence allowing different GPU threads to work on different signals. The advantages of the multi-signal approach are shown through several benchmarks involving the Self-Organizing Adaptive Map (SOAM) algorithm as a basis for evaluation. Having a graph-based termination condition that depends on the features of the network being grown, the SOAM algorithm allows assessing both functional equivalence and performances of the paradigm proposed without relying on arbitrary thresholds. Nonetheless, the evaluation proposed has a broader scope since it refers to a unified framework for the GPU parallelization of a generic SONN.
Integration of growing self-organizing map and bee colony optimization algorithm for part clustering
2018, Computers and Industrial Engineering
This study proposes a two-stage clustering algorithm for part clustering based on a self-organizing map (SOM) algorithm. The SOM algorithm has been applied widely in clustering since it can represent data pattern in two-dimensional topology. However, SOM algorithm requires a predetermined topology shape and connecting weights within the topology. Both of them can seriously influence the clustering performance. The proposed algorithm involves growing self-organizing map (GSOM) algorithm and bee colony optimization based self-organizing map (BCOSOM) to determine the SOM topology and improve the performance of SOM. In the first stage, GSOM is applied to determine the SOM topology. It is then followed by BCOSOM to fine-tune the SOM weights. The proposed GSOM + BCOSOM algorithm is compared with other algorithms, namely, particle swarm optimization (PSO), BCO, SOM, GSOM + PSOSOM, SOM + PSO, and SOM + BCO algorithms, using seven benchmark data sets. The computational results indicate that GSOM + BCOSOM algorithm is able to find a better solution than other algorithms. Furthermore, the proposed algorithm is also applied to a real-world problem about clustering components into part families for a medical furniture manufacturer in Indonesia. By clustering the components, the manufacturer can reduce production lead-time, setup time, production time, and work-in-process inventory in order to enhance the manufacturing efficiency. According to the GSOM + BCOSOM result, the components are grouped into three clusters. This study also identifies the corresponding characteristics of each cluster.
A hybrid metaheuristic and kernel intuitionistic fuzzy c-means algorithm for cluster analysis
2018, Applied Soft Computing Journal
Citation Excerpt :
Table 2 lists the tested parameter settings in this study. These values are taken from previous studies [33–36]. 30 independent runs are conducted for each combination of parameter settings.
Cluster analysis is a very useful data mining approach. Although many clustering algorithms have been proposed, it is very difficult to find a clustering method which is suitable for all types of datasets. This study proposes an evolutionary-based clustering algorithm which combines a metaheuristic with a kernel intuitionistic fuzzy c-means (KIFCM) algorithm. The KIFCM algorithm improves the fuzzy c-means (FCM) algorithm by employing an intuitionistic fuzzy set and a kernel function. According to previous studies, the KIFCM algorithm is a promising algorithm. However, it still has a weakness due to its high sensitivity to initial centroids. Thus, this study overcomes this problem by using a metaheuristic algorithm to improve the KIFCM result. The metaheuristic can provide better initial centroids for the KIFCM algorithm. This study applies three metaheuristics, particle swarm optimization (PSO), genetic algorithm (GA) and artificial bee colony (ABC) algorithms. Though the hybrid method is not new, this is the first paper to combine metaheuristics and KIFCM. The proposed algorithms, PSO-KIFCM, GA-KIFCM and ABC-KIFCM algorithms are evaluated using six benchmark datasets. The results are compared with some other clustering algorithms, namely K-means, FCM, Kernel fuzzy c-means (KFCM) and KIFCM algorithms. The results prove that the proposed algorithms achieve better accuracy. Furthermore, the proposed algorithms are applied to solve a case study on customer segmentation. This case study is taken from franchise stores selling women's clothing in Taiwan. For this case study, the proposed algorithms also exhibit better cluster construction than other tested algorithms.
A directed batch growing approach to enhance the topology preservation of self-organizing map
2017, Applied Soft Computing Journal
Citation Excerpt :
The issue makes SOM unsuitable for online learning and handling non-stationary datasets. To work around this limitation, a dynamic variant of the SOM, growing SOM (GSOM) were proposed [10,11] which has been applied to many different fields [7,12–16]. Instead of being confined to a predetermined number of neurons, GSOM offers a flexible structure and requires less number of epochs compared to the original SOM which enable the ability to learn the nonlinear manifolds in high dimensional feature space [17–19].
The growing self-organizing map (GSOM) possesses effective capability to generate feature maps and visualizing high-dimensional data without pre-determining their size. Most of the proposed growing SOM algorithms use an incremental learning strategy. The conventional growing approach of GSOM is based on filling all available position around the candidate neuron which can decrease the topology preservation quality of the map due to the misconfiguration and twisting of the map which could be a consequence of unexpected network growth and improper neuron addition and weight initialization. To overcome this problem, in this paper we introduce a batch learning strategy for growing self-organizing maps called DBGSOM which direct the growing process based on the accumulative error around the candidate boundary neuron. In the proposed growing approach, just one new neuron is added around each candidate boundary neuron. The DBGSOM offers suitable mechanisms to find a proper growing positions and allocating initial weight vectors for the new neurons.
The potential of the DBGSOM was investigated with one synthetic dataset and six real-world benchmark datasets in terms of topology preservation and mapping quality. Experimental results showed that the proposed growing strategy provides an enhanced topology preserved map and reduces the susceptibility of twisting compared to GSOM. Furthermore, the proposed method has a better clustering ability than GSOM and SOM. According to the lower number of neurons generated by DBGSOM, it needs less time to learn the manifold of the data points compared to GSOM.
Clustering and visualization of failure modes using an evolving tree
2015, Expert Systems with Applications
Citation Excerpt :
It also provides a topological view of the underlying data structure (Kohonen et al., 1996; Kohonen, 2001; Vesanto & Alhoniemi, 2000). A number of enhanced SOM models have been proposed, e.g. growing SOM (GSOM) (Matharage, Alahakoon, Rajapakse, & Pin, 2011; Kuo, Wang, & Chen, 2012), growing hierarchical SOM (GHSOM) (Huang & Tsaih, 2012), and ETree (Pakkanen, Iivarinen, & Oja, 2004, 2006). These enhancements overcome two shortcomings of SOM, i.e., the requirement of a pre-defined map size before learning (Kohonen, 2001; Vesanto & Alhoniemi, 2000) and the long learning time when a large map size is initiated (Pakkanen et al., 2004, 2006).
Despite the popularity of Failure Mode and Effect Analysis (FMEA) in a wide range of industries, two well-known shortcomings are the complexity of the FMEA worksheet and its intricacy of use. To the best of our knowledge, the use of computation techniques for solving the aforementioned shortcomings is limited. As such, the idea of clustering and visualization pertaining to the failure modes in FMEA is proposed in this paper. A neural network visualization model with an incremental learning feature, i.e., the evolving tree (ETree), is adopted to allow the failure modes in FMEA to be clustered and visualized as a tree structure. In addition, the ideas of risk interval and risk ordering for different groups of failure modes are proposed to allow the failure modes to be ordered, analyzed, and evaluated in groups. The main advantages of the proposed method lie in its ability to transform failure modes in a complex FMEA worksheet to a tree structure for better visualization, while maintaining the risk evaluation and ordering features. It can be applied to the conventional FMEA methodology without requiring additional information or data. A real world case study in the edible bird nest industry in Sarawak (Borneo Island) is used to evaluate the usefulness of the proposed method. The experiments show that the failure modes in FMEA can be effectively visualized through the tree structure. A discussion with FMEA users engaged in the case study indicates that such visualization is helpful in comprehending and analyzing the respective failure modes, as compared with those in an FMEA table. The resulting tree structure, together with risk interval and risk ordering, provides a quick and easily understandable framework to elucidate important information from complex FMEA forms; therefore facilitating the decision-making tasks by FMEA users. The significance of this study is twofold, viz., the use of a computational visualization approach to tackling two well-known shortcomings of FMEA; and the use of ETree as an effective neural network learning paradigm to facilitate FMEA implementations. These findings aim to spearhead the potential adoption of FMEA as a useful and usable risk evaluation and management tool by the wider community.
Comparison of visualization of optimal clustering using self-organizing map and growing hierarchical self-organizing map in cellular manufacturing system
2014, Applied Soft Computing Journal
The present research deals with the cell formation problem (CFP) of cellular manufacturing system which is a NP-hard problem thus, the development of optimum machine-part cell formation algorithms has always been the primary attraction in the design of cellular manufacturing system. In this proposed work, the self-organizing map (SOM) approach has been used which is able to project data from a high-dimensional space to a low-dimensional space so it is considered a visualized approach for explaining a complicated CFP data set. However, for a large data set with a high dimensionality, a traditional flat SOM seems difficult to further explain the concepts inside the clusters. We propose one such possible solution for a large CFP data set by using the SOM in a hierarchical manner known as growing hierarchical self-organizing map (GHSOM). In the present work, the two novel contributions using GHSOM are: the choice of optimum architecture through the minimum pattern units extracted at layer 1 for the respective threshold values and selection. Furthermore, the experimental results clearly indicated that the machine-part visual clustering using GHSOM can be successfully applied in identifying a cohesive set of part family that is processed by a machine group. Computational experience specifically with the proposed GHSOM algorithm, on a set of 15 CFP problems from the literature, has shown that it performs remarkably well. The GHSOM algorithm obtained solutions that are at least as good as the ones found the literature. For 75% of the cell formation problems, the GHSOM algorithm improved the goodness of cell formation through GTE performance measure using SOM as well as best one from the literature, in some cases by as much as more than 12.81% (GTE). Thus, comparing the results of the experiment in this paper with the SOM and GHSOM using the paired t-test it has been revealed that the GHSOM approach performed better than the SOM approach so far the group technology efficiency (GTE) measures of performance of the goodness of cell formation is concerned.

View all citing articles on Scopus

View full text

Integration of growing self-organizing map and continuous genetic algorithm for grading lithium-ion battery cells

Abstract

Graphical abstract

Highlights

Introduction

Section snippets

Literature review

Methodology

Simulation results and analysis

Model evaluation results

Conclusions

Acknowledgements

European Journal of Operational Research

Expert Systems with Applications

Journal of the Franklin Institute

Neural Networks

Applied Soft Computing

Neural Network

Neural Networks

Neural Network

Neural Networks

Applied Mathematics and Computation

European Journal of Operational Research

Decision Support Systems

Expert Systems with Applications

Expert Systems with Applications

Computers in Industry

Engineering Applications of Artificial Intelligence

Neurocomputing

Journal of Food Engineering

Energy Policy

International Congress Series

Computers & Mathematics with Applications

Data clustering: a review

ACM Computer Survey

Data Mining: Concepts and Techniques

The self-organizing map

Proceedings of the IEEE

Dynamic self-organizing maps with controlled growth for knowledge discovery

IEEE Transactions on Neural Networks

Online auction customer segmentation using a neural network model

International Journal of Applied Science and Engineering