Dragan Aleksendrić

The operational reliability of electric buses depends heavily on the thermal stability and longevity of their battery packs. In urban public transit, highly dynamic driving cycles characterized by severe current surges during regenerative braking and prolonged constant-current charging phases induce massive localized thermal stress. Traditional battery management systems often fail to dynamically balance energy recovery with the physical constraints of cell matrices, leading to accelerated degradation and potential safety hazards. This paper proposes an integrated cyber–physical framework that merges intelligent control with advanced prognostic monitoring and predictive thermal management. The core of the control layer is a deep Q-network (DQN) agent designed for real-time smart charging and dynamic regenerative braking current limitation (smart derating). This control loop is tightly integrated with a long short-term memory (LSTM) network-based digital twin (DT) operating at the edge level to mirror complex electro-thermal dynamics and track the battery state of health (SOH). To ensure transparency and proactive thermal control, a nonlinear model predictive control (NMPC) strategy is implemented, utilizing SHapley Additive exPlanations (SHAP) to map core heat generation and dynamically adjust for internal thermal lag. The predictive NMPC-xAI loop maintained cell temperatures within a tight, optimal envelope, preventing delayed under-cooling and reducing peak temperature fluctuations.

Vehicle diagnostic systems are crucial for the normal operation of vehicles and their propulsion-related systems. Undetected unusual behaviour of such systems makes the vehicle diagnostic system unreliable. Current diagnostic systems, such as On-Board Diagnostics (OBD), are limited to monitoring only specific systems in order to make fault decision. However, various anomalies, including drastic performance drops, vehicle tampering, and changes in the driving environment, often go undetected during OBD system testing, validation, and inspection. This research presents a novel explainable OBD anomaly detection pipeline that is able to detect anomalies based only on OBD data snapshots during processes of OBD validation and inspection. The novel approach is implemented using combined dimension reduction and data clustering methodologies. First, the data is transformed into a latent space using t-distributed Stochastic Neighbor Embedding (t-SNE), where the general structure of the anomaly in the data can be exploited. Subsequently, clustering using Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is applied to group similar normal data and identify anomalous patterns. The novelty of the solution is further extended with a feedback loop that suggests the root cause of OBD signals for individual anomalies using explainable AI (XAI) methodology, in this case Shapley additive explanations (SHAP). The proposed concept was verified and evaluated using real OBD snapshots with synthetically generated anomalies in two scenarios with different engine status, engine off and on, with an achieved accuracy of 92.89% and 96.45% for anomaly detection, respectively. The majority of anomaly causes in the form of specific OBD signals from propulsion- and emission-related systems were successfully explained using SHAP.

Increasing dynamic performance and the general level of automation of commercial vehicles emphasize the issue of safety. Modern braking systems focus on sustaining vehicle stability, often degrading the brake performance. The major downgrades of the braking performance are nearly impossible to model using a classical mathematical approach, making them not feasible to use in real braking system controllers. In this paper, the use of combined Neural Networks and Fuzzy logic for the control of the braking system of a commercial vehicle while maximizing performance and sustaining stability is proposed. The control system comprises adhesion estimation, an inverse brake model, and a fuzzy logic controller to keep the system giving optimal control signals in various brake conditions while sustaining vehicle stability and steerability. The results based on a semi-trailer system reveal the success of the proposed AI-based braking system algorithm while braking under varying conditions. Graphical abstract

The curing process of thermosetting resins plays a key role on the final quality of the composite material components. Soft computing techniques proved to be an efficient method to control and optimize the curing process, replacing the conventional experimental and numerical approaches. In this paper artificial neural network (ANN) and fuzzy logic control (FLC) were implemented together to predict and control the temperature and degree of cure profile during the autoclave curing process. The obtained outcomes proved the capability of ANNs and FLC with respect to the hard computing methods.

Manufacturing parameters of brake friction materials have a crucial impact on its friction and wear behavior. The effects of manufacturing parameters on wear have been modeled aiming at the optimization of facing behaviors. The influences of the five most important manufacturing parameters, such as molding pressure, molding time, molding temperature, heat-treatment temperature, and heat-treatment time, have been investigated. Their effects have been investigated on different formulations of the friction material and under different brake interface temperatures. The neuro-genetic optimization model was developed to be able to predict and optimize the effects of manufacturing parameters on the brake friction material wear.