Abstract
This study presents a novel approach to cope with the human behaviour uncertainty during Human-Robot Collaboration (HRC) in dynamic and unstructured environments, such as agriculture, forestry, and construction. These challenging tasks, which often require excessive time, labour and are hazardous for humans, provide ample room for improvement through collaboration with robots. However, the integration of humans in-the-loop raises open challenges due to the uncertainty that comes with the ambiguous nature of human behaviour. Such uncertainty makes it difficult to represent high-level human behaviour based on low-level sensory input data. The proposed Fuzzy State-Long Short-Term Memory (FS-LSTM) approach addresses this challenge by fuzzifying ambiguous sensory data and developing a combined activity recognition and sequence modelling system using state machines and the LSTM deep learning method. The evaluation process compares the traditional LSTM approach with raw sensory data inputs, a Fuzzy-LSTM approach with fuzzified inputs, and the proposed FS-LSTM approach. The results show that the use of fuzzified inputs significantly improves accuracy compared to traditional LSTM, and, while the fuzzy state machine approach provides similar results than the fuzzy one, it offers the added benefits of ensuring feasible transitions between activities with improved computational efficiency.

Abstract
As the need for replacing fossil and other non-renewable energy sources with renewables becomes more critical and urgent, wind energy appears to be among the two or three best choices for the short and medium time frames. The dominance of wind energy as the first choice in many regions, leads to an increasing impact of wind power quality on the overall grid. Wind energy’s inherent intermittent nature, both in intensity and longevity, could be an impediment to its adoption unless utility operators have the tools to anticipate the impact and integrate wind resources seamlessly by increasing or reducing its contribution to the overall capacity of the grid. The wind forecasting science is well established and has been the subject of serious study in multiple fields such as fluid dynamics, statistical analysis and numerical simulation and modeling. With the renewed interest and dependence on wind as a major energy source, these efforts have increased exponentially. One of the areas that shows great promise in developing improved forecasting tools, is the category of “Biological Inspired Optimization Techniques. The study presented in this paper is the result of a study to survey and assess an array of forecasting models and algorithms. © MIR Labs, www.mirlabs.net/ijcisim/index.html

Abstract
As the world's population grows, there is a need to find new sources of energy that are more sustainable. Photovoltaic (PV) energy is one of the renewable energy sources (RES) expected to have the greatest margin for growth in the near future. Given their intermittency, RES bring uncertainty and instability to the management of the power system, therefore it is essential to predict their behavior for different time frames. This paper aims to find the most effective forecasting method for PV energy production that could be applied to different time frames. PV energy production is directly dependent on solar radiation and temperature. Several forecasting approaches are proposed in this paper. A multiple linear regression (MLR) model is proposed to predict the monthly energy production based on the climatic parameters of the previous year. Different approaches are proposed based on first predicting the temperature and radiation and then applying the PV mathematical models to predict the produced energy. Three methods are proposed to predict the climatic parameters: using the average values, the additive decomposition, or the Holt-Winters method. Comparing the errors of the four proposed forecasting methods, the best model is the Holt-Winters, which presents smaller errors for radiation, temperature, and produced energy. This method is close to additive decomposition.

Abstract
Myocardial infarction, known as heart attack, is one of the leading causes of world death. It occurs when blood heart flow is interrupted by part of coronary artery occlusion, causing the ischemic episode to last longer, creating a change in the patient’s ECG. In this work, a method was developed for predicting patients with MI through Frank 3-lead ECG extracted from Physionet’s PTB ECG Diagnostic Database and using instantaneous frequency and spectral entropy to extract features. Two neural networks were applied: Long Short-Term Memory and Bi-Long Short-Term Memory, obtaining a better result with the first one, with an accuracy of 78%. © 2023, ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.

Abstract
This work addresses the wind farm (WF) optimization layout considering several substations. It is given a set of wind turbines jointly with a set of substations, and the goal is to obtain the optimal design to minimize the infrastructure cost and the cost of electrical energy losses during the wind farm lifetime. The turbine set is partitioned into subsets to assign to each substation. The cable type and the connections to collect wind turbine-produced energy, forwarding to the corresponding substation, are selected in each subset. The technique proposed uses a genetic algorithm (GA) and an integer linear programming (ILP) model simultaneously. The GA creates a partition in the turbine set and assigns each of the obtained subsets to a substation to optimize a fitness function that corresponds to the minimum total cost of the WF layout. The fitness function evaluation requires solving an ILP model for each substation to determine the optimal cable connection layout. This methodology is applied to four onshore WFs. The obtained results show that the solution performance of the proposed approach reaches up to 0.17% of economic savings when compared to the clustering with ILP approach (an exact approach).

Abstract
Currently, distributed software systems have evolved at an unprecedented pace. Modern software-quality requirements are high and require significant staff support and effort. This study investigates the use of a supervised machine learning model, a Multi-Layer Perceptron (MLP), for anomaly detection in microservices. The study covers the creation of a microservices infrastructure, the development of a fault injection module that simulates application-level and service-level anomalies, the creation of a system monitoring dataset, and the creation and validation of the MLP model to detect anomalies. The results indicate that the MLP model effectively detects anomalies in both domains with higher accuracy, precision, recovery, and F1 score on the service-level anomaly dataset. The potential for more effective distributed system monitoring and management automation is highlighted in this study by focusing on service-level metrics such as service response times. This study provides valuable information about the effectiveness of supervised machine learning models in detecting anomalies across distributed software systems.