Mathematical models form the basis of automation and digitalization. Control and optimization of industrial processes are important for increasing productivity and efficiency, especially in the sugar industry. This research focuses on modeling and controlling the juice extraction process, which is an important activity in sugar production. The mathematical model is obtained by creating a variable based on simple equations where the cane level in the Donnelly channel is the input and the juice output. The model captures the complexity of the process and provides a solid basis for the design of control systems. Two advanced control concepts: H-infinity control and model control (MPC) were used in MATLAB to meet the criteria. While H-infinity control provides performance in the presence of uncertainty and disturbances, MPC optimizes control performance by predicting future results. This paper observes and compares the results of two control systems to analyze their performance. This comparison highlights the advantages and limitations of each method. The research results are of great importance for increasing the efficiency and reliability of industrial processes in the sugar industry.

The energy efficiency, consistency of product quality, and optimization of operations for sugar manufacturing processes are some of the significant challenges faced by the sugar industry today due to the nonlinear dynamics of these processes. This paper describes how an innovative Digital Twin framework built using Reinforcement Learning (RL) for stepwise learning and adaptation to process control can be developed. First-principles Mathematical Models (FMMs) for four (4) critical Operational Units (OU) in the sugar production process: Cane Maceration, Juice Clarification, Vacuum Evaporation and Crystallization of Sugar are developed. The individual FMMs were combined to form a Digital Twin that continually synchronizes and updates with the respective physical systems using state estimation and Sensor Fusion (SF). The Proximal Policy Optimization (PPO) RL agent developed for this paper learns to control these operations through Multi-Objective Control Policies (MOCP) that optimize Sugar Yield, Energy Consumption, Product Purity, and Operational Safety. Results from simulations demonstrate a 12.0% decrease (4.2 -> 3.7 MJ kg-1) in Specific Energy Consumption, 3.1% increase (87.5 -> 90.2%) in Yield, 3.6% increase (88.3 -> 91.5%) in Pure Water, and 57.1% decrease in Process Instability over that of traditional PID Controllers. The authors built a three-tiered safety enforcement architecture to ensure that all policies learned by the PPO RL agent are implemented correctly. Through an economic analysis of a typical 25,000 T/season Sugar Factory (Indian Rupee89.0 Million/year), it was found that the payback period for this investment is 1.3 months after installation of the RL Agent. By integrating the Physical Principles of a Sugar Manufacturing Process with Data-Driven Optimizations and learning, this research provides a scalable Digital Twin Framework for the development of next-generation sugar factories.

The paper has played a vital role in the life of humans from ancient times covering a vast range of applications such as writing, packaging, and printing. The present paper is presenting a comprehensive review of various optimization and control methodologies, ranging from conventional to advanced ones, pertaining to the paper mill. The final goal of these control strategies is to upgrade the mill’s production and quality in presence of multiple technical challenges such as nonlinear and multivariable nature of the involved processes, various disturbance parameters, and time delays. In this work, the integration of machine learning with paper mill process is illustrated. For any manufacturing process, the final product quality is the key goal. There are various traditional techniques which have already been practiced for final produced paper quality in paper mills. This paper highlights the capability of support vector machine (SVM) algorithm to assess the produced paper quality, capturing the two crucial inputs viz. the pulp consistency and the headbox level. The basic goal of this research is twofold, firstly it presents an exhaustive literature survey exploring various strategies which are practiced currently in the domain of control and optimization of various paper mill processes. Secondly, it intends to develop and evaluate various SVM and SVM-RF hybrid models using MATLAB for assessment of quality of final product on basis of two parameters- pulp consistency and head box level. Finally, genetic algorithm has been employed in MATLAB for multivariate optimization.

The intensive development of computer networks has made the necessity of effective and intelligent resource allocation systems extremely high. Conventional optimization techniques very often fail to deal with the nonlinear relation and multiobjective trade-offs between the network resources including bandwidth, CPU allocation, memory usage, power consumption, and buffer capacity. To solve these issues, this paper outlines a Differential Evolution (DE)-based resource optimization model to computer networks. The model designed uses MATLAB to allow the dynamic interaction of network parameters and the allocation to maximize throughput and minimize the operational cost. The algorithm will make repeated adjustments to a population of solutions by using mutation, crossover and selection to arrive at an optimal resource configuration. The analysis of the simulation reveals that this algorithm provides a balanced use of the resources, and an enhanced optimization efficiency. MATLAB has been used in the simulation, visualization and quantitative assessment of the algorithm performance.

This research article investigates the implementation of Cuckoo Search (CS) optimization technique to enhance intrusion detection systems (IDS) of cybersecurity using MA TLAB. Detection of intrusion plays a crucial role in finding and mitigating unpermitted access attempts in networks. The efficacy of an IDS is defined majorly by two metrics: detection accuracy, which means network's ability to identify correct threats, and other one is false positive rate, which includes cases of mistakenly identified as benign activities. In this research work, attempt has been made to optimize two key parameters of IDS viz. decision threshold and model weight, to improve the performance. The CS algorithm, a heuristic and bio-inspired optimization method, is applied as it is capable of exploring complex search space efficiently. By employing probabilistic abandonment and levy flights mechanism, it bypasses local optima, hence exhibits improved global search efficacy. Further the impact of adjusting the abandonment probability (Pa) on the performance of algorithm, monitoring the progression of fitness values over several iterations, has been investigated. The fitness function of IDS is articulated to deliver maximum detection accuracy at the same time minimization of false positive rate, crafting a sturdy intrusion detection mechanism. Experimental outcomes exhibit the adaptation of CS algorithm in optimization of IDS attributes, making it a robust approach for improving cybersecurity systems' detecting capabilities in real-time applications.

It is important to optimise Battery Management Systems (BMS) to improve the safety, lifetime, and performance of EV batteries. The purpose of this work is to maximize two most important parameters of the battery viz. SOC and SOH using GA (Genetic Algorithm) in MATLAB. This has been optimized using analysis and simulations that have been conducted to present the performance improvement of the BMS by this optimization technique in the presence of different critical constraints. To begin with, a mathematical modeling has been provided followed by the maximization of SOC and SOH of battery of an EV using GA. The result of SOC and SOH optimization of GA also gives the values of some important parameters including voltage, temperature, capacity, impedance, and maximum fitness.

This research article demonstrates the optimum load allocation in a hybrid wireless network including Wi-fi, cellular and satellite networks, so that congestion is minimized, using Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) in MATLAB. The model focusses on reducing congestion in network through optimization in distribution of traffic load in the three networks. Congestion weights for every network state its importance and priority in minimizing congestion for each domain. The PSO technique starts with initializing swarm particles, updating position and velocity repeatedly in order to converge to an optimal distribution of traffic load to minimize congestion. PSO's fitness function evaluates every particle's position, so it delivers realistic lower bounds and meets required capacity constraints of the network. Parallelly, ACO technique is implemented which mimics the ants' behavior wherein every ant creates a solution that uses heuristic functions and pheromone levels navigating them in direction of optimized load allocation for network. The pheromone and evaporation parameters are tuned to balance exploitation and exploration of the solution space, to converge at effective solutions. The comparative study states the strength and weakness of considered optimization strategies to achieve a balance in congestion in hybrid networks, delivering useful insights for engineers targeting improved operational efficiency and quality of service.

In modern digital security, predicting cyber security risk is a critical aspect. This research is presenting a comparative analysis of two distinct AI (Artificial Intelligence) based models viz. Fuzzy Inference System and Adaptive Neuro Fuzzy Inference System for predicting cyber security risk on the basis of 3 parameters namely authentication strength, traffic anomaly and network load. Both the models are developed in MATLAB, for which a data set of 100 samples is utilized, 85% for training and 15% for testing. Gaussian membership functions have been formed for all input and output variables. The evaluation of this system is done comparing the anticipated risk values with actual values. The ANFIS model has been initialized with grid partition fuzzy inference system. It is trained in 200 epochs considering a small tolerance for ensuring convergence. This trained model is hence tested on the unseen inputs for predicting cyber security risk. Its output has been compared with that of fuzzy model. Evaluation metric namely root mean square error (RMSE) has been used for assessing the prediction accuracy. It has been found that the ANFIS is outperforming the conventional FIS approach displaying a smaller R.M.S.E. and enhancing prediction accuracy. This study outlines the advantages of integrating neuro adaptive learning into FIS for assessing cyber security risk, upgrading risk prediction accuracy. Therefore, ANFIS comes out to be more reliable technique pertaining to cyber security application.

BLDC (Brushless DC) motors are well-known for their high reliability and smaller maintenance cost. For attaining optimal performance, accurate speed control is critical, especially in dynamic load conditions. This work proposes the designing of a neural-network based PI (Proportional-Integral) controller developed in MATLAB for improving the control performance of BLDC motor. Mathematical modelling approach has been employed for describing the dynamics of motor including current, voltage, back emf, and torque equations. The training data containing control voltages and speed errors for feedforward neural network is generated on the basis of a conventional PI controller used initially. It enables the dynamic adjustment of the gains of PI controller in real-time. A neural network is integrated in the proposed controller to overcome the shortcomings of conventional PI controller like fixed gain parameters and poor adaptability to fluctuating operating conditions. Simulations have been conducted for evaluating the performance of developed controller under various conditions. The results exhibit remarkable improvements in key performance parameters.

Dynamic spectrum allocation (DSA) refers to attaining efficient utilization of spectrum according to the fluctuating network conditions. This work is presenting Fuzzy Logic based DSA with a goal of addressing the challenges caused by fluctuating interference levels, traffic load, and user demands. An FIS (Fuzzy Inference System) has been designed in MATLAB to achieve this goal. The rules for this FIS have been formulated to take care of various network conditions, guaranteeing optimal allocation in a variety of conditions. The developed fuzzy inference system is validated comparing the simulated output with the actual spectrum values. The key metrices in terms of which the performance is evaluated are MAE, MSE, RMSE, and R2. The final results exhibited an encouraging predictive accuracy. This research outlines the robustness and interpretability of fuzzy logic in the domain of spectrum management in wireless communication.

Because of rapid development of technology and escalating consumer demand for electronic devices, e-waste (Electronic Waste) has turned out to be a severe environmental concern. An efficient management of e-waste is crucial for minimizing environmental hazards and optimizing resource recovery. The present research is applying GA (Genetic Algorithm) and ACO (Ant Colony Optimization) methods, in MATLAB, for optimizing the e-waste collection points allocation to the recycling centers. The factors like e-waste amounts, transportation cost, and center capacities have been considered for determining an optimal distribution strategy. The performance of GA and ACO in minimizing the overall cost has been evaluated. The results revealed that both techniques are optimizing the allocation process effectively. However, ACO exhibits faster convergence, and GA provides more diverse solutions. This research contributes to development of sustainable and efficient e-waste management approaches by integrating AI-driven optimization techniques.

With the speedy advancements in EVs (Electric Vehicles) and their incorporations into power grid, the V2G (Vehicle-To-Grid) concept has emerged as an encouraging solution for cost-effective utilization of energy, and grid stability. This research is presenting an optimization method for EV charging using fuzzy logic-based method aiming on reduction of cost and enhancement in battery life. A fuzzy inference system (Mamdani type) has been developed, using MATLAB, with three input parameters viz. electricity price, SOC (state of charge), and grid demand for determining an optimal charging decision. The fuzzy rules and membership functions have been formulated systematically for ensuring and intelligent charging scheme which balances the grid demand and economic benefits. The developed model has been evaluated utilizing 20 test cases based on a secondary data set pertaining to real world fluctuations in charging conditions. The results indicated that the developed fuzzy model exhibits noteworthy reduction in electricity expenses and an extended battery life by reducing the charging cycles appreciably. The findings of this research underscore the efficacy of fuzzy Logic in optimization of EV charging, guaranteeing economic advantages for users, along with maintaining the grid stability.

The idea of this research is to design an efficient control system for a two tank system using two advanced control approaches – IMC (internal model control) and MPC (model predictive control). The precise control of the two-tank system is crucial for desired performance and efficiency. The paper begins with an overview of the two-tank system, highlighting its dynamic behavior and the challenges in regulating simultaneously the liquid levels in both the tanks. It underlines the need for advanced control strategies for enhancing the stability and responsiveness of the system. The first part of the study presents the development of IMC controller which handles the complex dynamics and mitigates the disturbances promisingly. Finally, the MPC technique is explored for the control of considered two-tank system. The MPC uses a prediction based approach which considers the future behavior of system in decision making which makes it well-suited to handle nonlinearities and constraints. The performance of both the established controllers is also compared. All simulation work is carried out in MATLAB.

This research is addressing a very critical parameter Air Quality Index (AQI) having a severe impact on human health. The management of AQI is of utmost importance in existing environmental discourse, demanding reliable and accurate prediction models for AQI. This research is contributing considerably in this area by exploring two different approaches of AQI prediction modeling. One is the classical approach which is known for its simplicity that is Linear Regression model. Another one is a more advanced technique, integrating fuzzy logic & neural networks, named Adaptive Neuro-Fuzzy Inference System (ANFIS) model. MATLAB has been used as a primary tool for developing and analyzing the models.

This research work focuses on the development and comparison of soil fertility classification models based on Ensemble methods. This study utilizes machine learning based algorithms for analyzing and assessing the soil fertility levels, a vital feature in sustainable agriculture. The models are developed using three different Ensemble methods in MATLAB. The ensemble learning approach has ability of handling complex datasets. It provides very accurate anticipations. The research considers three factors viz. pH, Nitrogen concentration, and Phosphorous concentration for soil fertility classification. This work offers valuable visions into effectiveness of Ensemble techniques in soil fertility classification, providing a practical approach to farmers & agriculture practitioners for making informed decisions pertaining to soil management practices.

In the field of sugar production, guaranteeing consistent quality is vital for both consumers and industries. The goal of this research is to establish models, based on machine learning, capable of classifying the produced sugar quality into 3 distinct classes viz. High, Low, and Medium. There are several parameters associated with the complex production process es of sugar Mill. The developed models will include the potential of machine learning techniques to speedily and precisely classify the sugar quality. To accomplish this, a secondary dataset based on numerous runs of sugar production has been used here to train and test two machine learning based models using the Decision Tree, and K-Means clustering techniques in MATLAB. The classification accuracy of both models has also been Investigated, The successful creation of machine learning based models for sugar quality classification within the sugar mill has substantial implications. By automation of classification process, impending defects or inconsistencies in sugar quality can be swiftly identified, enabling timely interventions for oprlmizing production process. This research provides noteworthy potential for overall sugar quality improvement, improving customer satisfaction, and reinforcing the competitivenes s of sugar industries in the market.

Paint & Coating industries have displayed a significant role in a variety of areas e.g. aerospace, construction, and automotive applications. However, a big challenge that has been associated with this process is the precise control of coating process for ensuring high-quality finishes maintaining consistent uniformity and thickness. The conventional control strategies can’t give promising results due to uncertainties and nonlinearities existing in these processes. This research is exploring the potential of H infinity control in enhancing the performance and robustness of control system for a considered second-order paint & coating process model. For this model, a conventional PID controller, and a H-infinity based controller have been designed using MATLAB. The improvement in the control performance by H -infinity technique has been highlighted by comparing it with the conventional PID control system in terms of various time domain and frequency domain parameters.

The emergence of 5G communication networks has revolutionized unrivalled demands for dynamically allocated spectrum to handle the sharp increase of services and devices. Effective management of spectrum is crucial to maximize throughput, minimize interference, and confirming expected quality of Service (QoS) for various applications. This article investigates the DSA (Dynamic Spectrum Allocation) optimization in 5G communication networks by utilizing two evolutionary algorithms namely Genetic Algorithms (GA) and Particle Swarm Optimization (PSO). The DSA optimization task has been developed as a multi-objective optimization problem, focusing to improve performance of network while referring to major challenges implicit in 5G communication networks, such as heterogeneous configuration of network, interference management, strict QoS requirements, and energy efficacy. The objectives of optimization comprise of minimum interference, maximum throughput, and accomplishment of various QoS requirements of its users.Rigorous simulations were performed in MATLAB to implement and estimate the performance of GA and PSO in the given 5G communication network scenarios. Results depict that both the algorithms majorly improve network performance and spectrum efficacy, PSO exhibiting fast convergence and GA delivering better diverse solutions. Comparatively, strengths and bottlenecks of both the approaches, offers depth into its application to 5G networks.

In wireless communication, the Signal to Noise Ratio of the received signal plays an instrumental role in deciding the quality of reception. So, for network optimization, its accurate estimation is paramount. The prime goal of this work is predictive modelling pertaining to SNR in 5G communication using Machine Learning (ML) algorithms. Two different ML techniques which have been employed in this work using MATLAB are Gaussian Process Regression (GPR) and Least Squares Boosting (LSBoost). The GPR is a robust technique based on probabilistic predictions, on the other LSBoost belongs to the family of ensemble learning techniques. As there are several parameters on which the SNR of a 5G system depends, the corresponding dataset is highly non-linear and complex. So, the conventional techniques of estimation can’t give promising results in this case. This research elaborates the application of machine learning algorithms in making accurate predictions in such cases. The findings of this work could be effectively employed in optimization and resource allocation in 5G networks.

This research is based on ML (Machine Learning) based predictive modelling for assessing the SOC (State of Charge) of a BMS (Battery Management System). For the reliable and efficient use of a battery -based system, mainly in EVs, the accurate estimation of SOC is paramount. In this work, a secondary dataset including 93 samples of 4 key parameters namely Voltage, Current, Temperature, and SOC has been used to train the ML models for SOC estimation. The ML techniques which are implemented here in MATLAB are Linear Regression, and Support Vector Machine (SVM), for SOC prediction. 80 % of this dataset has been used for training and 20 % for testing. The Linear regression model provides a basic linear approach for SOC estimation. Whereas, the SVM has the ability of handling non-linear connection between input parameters and the output. The performance is compared on the basis of the test data inputs. Various graphical representations are also shown for illustrating the performance of developed models. The findings contribute significantly towards advancements in intelligent BMSs.