Yoshikazu Fukuyama

This paper proposes a kernel principal component analysis (KPCA) based multivariate statistical process control (KPCA-MSPC) method for fault detection of refrigeration showcase systems using a feature selection method with maximal information coefficient (MIC). Refrigeration showcase system data include non-linear relationships among pairs of features, and only normal data can be available for training generally. KPCA-MSPC is suitable for the fault detection because it is an unsupervised method and can handle non-linear relationships. In showcase systems, a large number of measured data can be obtained and they can be utilized as features for fault detection. However, considering system costs, the number of sensors installed in the showcase systems and the amount of data stored in data centers are limited. Therefore, a feature selection method based on MIC and k-nearest neighbor algorithm (KNN) (MIC-KNN-FS) suitable for KPCA-MSPC is proposed. The effectiveness of the combination of KPCA-MSPC and the proposed MIC-KNN-FS for showcase systems is verified by comparison with the Laplacian Score feature selection method (LS-FS) and the KNN feature selection method (KNN-FS), which are typically utilized as feature selection methods, and cumulative autoencoders (CAE) and MSPC based on PCA (PCA-MSPC), which are unsupervised fault detection methods.

This paper proposes a fault detection method for hydroelectric generators by robust random cut forest (RRCF) with feature selection using Hilbert-Schmidt Independence Criterion (HSIC). It is difficult to obtain fault data because faults unlikely occur in hydroelectric generators. Therefore, only normal data can be utilized for developing a fault detection method. Since data of hydroelectric generators have non-linear correlation, the fault detection method must consider the correlation. In this paper, the fault detection method of hydroelectric generators is assumed to be installed as cloud services in a data center. Therefore, there are two requirements for the applied method in order to develop a practical fault detection system. The first one is that the method has to generate an appropriate fault detection model for each hydroelectric generator in a short time in order to reduce CPU cost in the data center. The second one is that the limited number of features must be selected in order to save storage cost in the data center and communication cost through the internet. Therefore, if all data of hydroelectric generators are gathered, communication and CPU usage fees must be expensive. In order to reduce the costs, it is required to utilize only data with effective features for fault detection. The proposed method can construct models using only normal data, detect faults using data with non-linear correlation, and select effective features for fault detection of hydroelectric generators. The proposed method is compared with an isolation forest (IF) based method with all features, an IF based method with features of the highest accuracy, a RRCF based method with all features, and a RRCF based method with features of the highest accuracy. Area Under the Curve (AUC) values and calculation time are evaluated from the point of view of practical application of fault detection in data center. The proposed method is verified to be the most effective from the point of view. The results are verified using the Mann-Whitney U test.

This paper proposes improved reactive tabu search for optimal vehicle routing of vehicles considering allocation of secondary depots in case of natural disasters. In recent years, many large-scale disasters have occurred in Japan and large-scale disasters may occur in the near future, unfortunately. In case of large-scale disasters, emergency relief supply logistics is an important issue since it will affect lives of evacuees. The proposed method modifies a stored method of a tabu list, introduces aspiration criteria, and applies a searched list using a hash function for simultaneously solving the optimal number of secondary depots, the optimal placement and the distribution areas of secondary depots, and the optimal routes of vehicles in the distribution area of secondary depots. It is verified that the proposed method can generate higher quality optimal vehicle routing and reduce computation time than the conventional reactive tabu search based method with actual evacuation facility data of Chiba City, which are planned to be used in case of natural disasters. The results are confirmed by Wilcoxon test.

This paper proposes a fault detection method for refrigerated showcase by Pasting based Artificial Neural Networks (PANNs) using Parallel Multi-population Modified Brain Storm optimization (PMP-MBSO) and Correntropy. PMP-MBSO is an evolutionary computation technique and Correntropy is one of loss functions (LFs) for ANN. At present, in Japan, the number of convenience stores is about 50,000, and refrigerated showcases placed in each store have various characteristics. Since it is difficult to adjust fault detection methods for all different refrigerated showcases, an automatic parameter adjustment method for refrigerated showcase systems is required. The proposed PANNs using PMP-MBSO and Correntropy can realize the automatic adjustment. Using actual operation data of showcases, detection accuracy of the proposed method is verified to be higher than those of the ANNs using least square error (LSE) and stochastic gradient descent (SGD), and the ANNs using Correntropy and MBSO. Effectiveness of the proposed method is verified by applying Friedman test. It is verified to speed up the proposed method as well.

This study proposes improved parallel reactive hybrid particle swarm optimization (IPRHPSO) using an improved neighborhood schedule generation method for the integrated framework of optimal production scheduling and operational planning of an energy plant in a factory. Conventionally, in an energy plant, fixed loads of various tertiary energies have been utilized to solve the optimal operational planning of an energy plant so far. Additionally, in production lines, only the minimization of production time has been yet considered. Therefore, the secondary energy cost of a factory cannot be reduced accurately. However, in this study, the loads of various tertiary energies are calculated according to the candidates of production scheduling and the optimal operational planning of an energy plant is determined using the tertiary energies. This can explicitly reduce the secondary energy cost of a factory. The proposed method was applied to 10 jobs and machine JSPs each. Accordingly, it was verified that it can minimize the secondary energy cost and production time simultaneously, and realize fast computation through parallel computation using IPRHPSO.

This paper proposes a fault detection (FD) method for refrigerated showcase by a Correntropy based Artificial Neural Network (CANN) using Modified Brain Storm Optimization (MBSO), which is an evolutionary computation technique. Since over 55,000 convenience stores exist in Japan, developing FD models for each showcase with various features is difficult for experts. For the purpose of solving this challenge, an automatic ANN parameter training method for a variety of showcase systems should be developed. The proposed CANN using MBSO is verified to be more accurate than the conventional least square error (LSE) based ANNs (LANNs) using stochastic gradient descent (SGD) and the conventional CANNs using fast brain storm optimization (FBSO) with actual measurement data of showcase.

The authors wish to make the following corrections to their paper

This paper proposes dependable multi‐population improved brain storm optimization with differential evolution for optimal operational planning of energy plants. The problem can be formulated as a mixed‐integer nonlinear programming problem and various evolutionary computation techniques such as particle swarm optimization (PSO), differential evolutionary PSO (DEEPSO), multi‐population DEEPSO (MP‐DEEPSO), and brain storm optimization have been applied so far. When optimal operational planning of numbers of energy plants is calculated simultaneously in a data center, a challenge is to generate optimal operational planning as rapidly as possible considering control intervals and numbers of treated plants. One of the solutions for the challenge is speeding up by parallel and distributed computing. It utilizes numbers of processes and countermeasures for various faults of the distributed processes should be considered. Moreover, successive calculation at every control interval is required for keeping customer services. Therefore, sustainable (dependable) calculation keeping appropriate solution quality is required even if some of the calculation results cannot be returned from distributed processes. It is verified that total energy cost by the proposed dependable multi‐population improved brain storm optimization with differential evolution strategy based method is lower than those by the compared methods, and higher quality of solutions can be kept even with high fault probabilities.

This paper proposes total optimization of energy networks in a smart city by multi-population global-best modified brain storm optimization (MP-GMBSO). Efficient utilization of energy is necessary for reduction of CO₂ emission, and smart city demonstration projects have been conducted around the world in order to reduce total energies and the amount of CO₂ emission. The problem can be formulated as a mixed integer nonlinear programming (MINLP) problem and various evolutionary computation techniques such as particle swarm optimization (PSO), differential evolution (DE), Differential Evolutionary Particle Swarm Optimization (DEEPSO), Brain Storm Optimization (BSO), Modified BSO (MBSO), Global-best BSO (BSO), and Global-best Modified Brain Storm Optimization (GMBSO) have been applied to the problem. However, there is still room for improving solution quality. Multi-population based evolutionary computation methods have been verified to improve solution quality and the approach has a possibility for improving solution quality. The proposed MS-GMBSO utilizes only migration for multi-population models instead of abest, which is the best individual among all of sub-populations so far, and both migration and abest. Various multi-population models, migration topologies, migration policies, and the number of sub-populations are also investigated. It is verified that the proposed MP-GMBSO based method with ring topology, the W-B policy, and 320 individuals is the most effective among all of multi-population parameters.