Advanced machine learning algorithms are increasingly utilized to provide data-based prediction and decision-making support in Industry 4.0. However, the prediction accuracy achieved by the existing models is insufficient to warrant practical implementation in real-world applications. This is because not all features present in real-world datasets possess a direct relevance to the predictive analysis being conducted. Consequently, the careful incorporation of select features has the potential to yield a substantial positive impact on the outcome. To address the research gap, this paper proposes a novel hybrid framework that combines the feature importance detector - local interpretable model-agnostic explanations (LIME) and the feature interaction detector - neural interaction detection (NID), to improve prediction accuracy. By applying the proposed framework, unnecessary features can be eliminated, and interactions are encoded to generate a more conducive dataset for predictive purposes. Subsequently, the proposed model is deployed to refine the prediction of electricity consumption in foundry processing. The experimental outcomes reveal an augmentation of up to 9.56% in the R2 score, and a diminution of up to 24.05% in the root mean square error.

Photosynthesis is a crucial biochemical process for plant growth, which is determined by multiple environmental factors and other organic matter. In the horticultural industry, the environmental conditions in commercial greenhouses directly impact the quality of productions. Predicting the Net Photosynthesis (Pn) of plants based on the environmental parameters can help growers optimize the climate in greenhouse systems, thereby ensuring the quality of production. Meanwhile, due to the greenhouse climate can be controlled according to the prediction results, excess energy consumption can be avoided, so the production cost can be reduced. However, since the photosynthesis reaction is a highly nonlinear biochemical process, it is difficult for traditional algorithms to describe the hidden effects of individual elements. In previous related works, polynomial fitting was utilized for modeling the relation between Pn and environmental elements. In this paper, a Deep Learning (DL) method is explored to predict the Pn based on three inputs: light level, CO2 concentration and temperature. An exponential decay learning rate is applied in the training process to ensure convergence performance while increasing the convergence speed. Then, the performance of various Deep Neural Network (DNN) architectures is experimented and compared within this modeling problem. Finally, through a comprehensive analysis of the accuracy of individual architecture, a particular architecture is determined to solve this problem. According to the test results, the selected DNN can successfully predict the Pn based on the three environmental elements with high accuracy.

Security of energy supply and energy efficiency are challenges for many university campuses. Campus microgrid with distributed energy resources is considered as a solution for solving these challenges. However, the microgrid adoption for university campuses has to consider various factors due to the cost of DER installation and operation, campus activities and energy regulation etc. This paper investigates the possibility and feasibility of microgrid adoption for a university in Denmark, by analyzing the electricity and heat consumption patterns, the capacity of existing installed DERs, and the benefits and barriers for future installation of possible DER technologies. Based on this analysis the paper proposes three energy scenarios with different potential DER solutions, and provides simulations for investigating the energy balance and economic feasibility of each scenario.

Over the last decades, evolutionary algorithms have been extensively used to solve multi-objective optimization problems. However, the number of required function evaluations is not determined by nature of these algorithms which is often seen as a drawback. Therefore, a robust and reliable termination criterion is needed to stop the algorithm. There is a huge amount of knowledge encapsulated in the studies targeting termination criteria in evolutionary algorithms, but an updated integrated overview of this knowledge is missing. For this reason, we aim to conduct a systematic research through a comprehensive literature study. We extended the basic categorization of termination criteria to a more advanced one that takes the most common used termination criteria into consideration based on their specifications and the way they have been evolved over time. The survey is concluded by suggesting a road-map for future research directions.

Non-trivial real world decision-making processes usually involve multiple parties having potentially conflicting interests over a set of issues. State-of-the-art multi-objective evolutionary algorithms (MOEA) are well known to solve this class of complex real-world problems. In this paper, we compare the performance of state-of-the-art multi-objective evolutionary algorithms to solve a non-linear multi-objective multi-issue optimization problem found in Greenhouse climate control [1]. The chosen algorithms in the study includes NSGAII, ε-NSGAII, ε-MOEA, PAES, PESAII and SPEAII. The performance of all aforementioned algorithms is assessed and compared using performance indicators to evaluate proximity, diversity and consistency. Our insights to this comparative study enhanced our understanding of MOEAs performance in order to solve a non-linear complex climate control problem. The empirical findings of this comparative study show that based on the performance indicators, three algorithms, ε-MOEA, ε-NSGAII and NSGAII outperform the other algorithms and provide high quality solution sets in an appropriate time.

Any non-trivial software system has to be upgraded regularly to incorporate bug fixes and security patches or simply to keep up with the inevitable evolution in end-user requirements. Software upgrading is challenging, especially when it comes to online upgrading of running systems. In this paper, we present the current status of Gosh!, a dynamic-software-updating system for Java, which provides comprehensive support for changing class definitions of live objects, including adding, removing and moving fields, methods, classes and interfaces anywhere in the inheritance hierarchy. Prior to the acquisition by zeroturnaround.com, Gosh! was known as Javeleon. In this paper we demonstrate the capabilities of Gosh! by performing a dynamic updating experiment on five consecutive revisions of the classical arcade game Breakout. Based on the result of this experiment we show that dynamic updating of class definitions for live objects may under some circumstances result in different run-time behavior than would be observed after a cold restart of the upgraded application. Finally, we conclude by discussing the implication of this finding for future research directions within dynamic software updating.

Feature-centric comprehension of source code is necessary for incorporating user-requested modifications during software evolution and maintenance. However, such comprehension is difficult to achieve in case of large object-oriented programs due to the size, complexity, and implicit character of mappings between features and source code. To support programmers in overcoming these difficulties, we present a feature-centric analysis tool, Featureous. Our tool extends the NetBeans IDE with mechanisms for efficient location of feature implementations in legacy source code, and an extensive analysis of the discovered feature-code relations through a number of analytical views.

Abstract The ability to dynamically change the behavior of an application is becoming an important issue in contemporary rich client software development. Not only can programmers benefit from dynamic updates during the development of concurrent applications where recreation of complex application states can be avoided during test and debugging but also at post‐deployment time where applications can be updated transparently without going through the well‐known halt, redeploy and restart scheme. In this paper, we explain how our dynamic update framework achieves transparent dynamic updates of running Java applications while guaranteeing both type and thread safety. A novel feature of our approach is that it supports full redefinition of classes by allowing changes to the type hierarchy. Our approach is based on a lightweight runtime system, which is injected into an application via bytecode transformations at class loading. We show how our approach can add dynamic update capabilities to rich client development by integrating it with the NetBeans rich client platform. Performance experiments on our NetBeans implementation show that the overhead of our approach is low when applied to component application programming interface classes. To the best of our knowledge no other existing approach achieves the same level of low overhead and programming transparency. Copyright © 2009 John Wiley & Sons, Ltd.

Support for dynamic and client-specific customization is required in many application areas. We present a (distributed) application as consisting of a minimal functional core, implemented as a component based system, and an unbound set of potential extensions that can be selectively integrated within this core functionality. An extension to this core may be a new service due to new requirements of end users. Another important category of extensions we consider are non-functional services such as authentication, which typically introduce interaction refinements at the application level. In accordance with the separation of concerns principle, each extension is implemented as a layer of mixin-like wrappers. Each wrapper incrementally adds behavior and state to a core component instance from the outside, without modifying the component's implementation. The novelty of this work is that the composition logic, responsible for integrating extensions into the core system, is externalized from the code of clients, core system and extensions. Clients (end users, system integrators) can customize this composition logic on a per collaboration basis by 'attaching' high-level interpretable extension identifiers to their interactions with the core system.