Developing a robust and efficient data-driven digital twin system for industrial thermal power systems remains challenging due to data drift, change in operating behaviour of the system and ineffective data-sampling issues for data-driven model development. We present data-efficient model training framework that incorporates data sampling from large volumes of asymmetric and controlled data regimes of industrial power systems. Artificial Neural Network (ANN) model is trained on the sampled and representative dataset to predict turbine heat rate (THR) of 660-MW capacity thermal power plant. Later, THR is minimized by a constrained non-linear optimisation technique at 50 %, 75 %, and 100 % capacity discharge of power plant, and the optimisation-based results are validated in the operation of the power plant with the mean absolute percentage errors of 0.79 %, 2.98 % and 0.33 % respectively. The analysis on cost of operation and carbon dioxide (CO2) reduction reveals that minimizing THR through the data-efficient model training and optimization framework can save around 13 million USD with a reduction of 28 kilotonnes (kt) of CO2 per year. Finally, the data-efficient trained ANN model is deployed as a digital twin system for monitoring the THR and is found to be more than 90 % accurate on 19000 min of real-time monitoring window. This research paves the way for data-efficient sampling from the controlled datasets of industrial systems that leads to improved generalisation capacity of the trained machine learning models for their integration in the digital twin systems for monitoring the performance of industrial power systems.

AI adoption in industrial power operations—including gas power plants—remains slow. Given the "black-box" nature of neural networks in safety-critical contexts, domain-enforced neural optimisation with operator-friendly interfaces is seldom explored for techno-enviro-economic gains. We present a domain-enforced, operator-in-the-loop neural simulation platform that embeds neural surrogate models within a nonlinear optimisation framework to estimate optimised operating levels of process variables, implemented for a 395 MW gas turbine system. We train surrogate models-Artificial Neural Network (ANN), Data-Information-integrated Neural Network (DINN), and Kolmogorov-Arnold Networks (KAN)—to predict Power (MW), Turbine Heat Rate (kJ/kWh), and Thermal Efficiency (%). On test data, KAN attains slightly higher predictive performance (R-squared of 0.96 or greater) than ANN (R-squared of 0.92 or greater) and DINN (R-squared of 0.93 or greater). The embedded models both predict and evaluate the likelihood of attaining target performance values given typed-in operating levels of process inputs. A Mahalanobis-distance constraint introduces the operator into the loop and enforces domain consistency while estimating energy-efficient, domain-consistent operating levels to deliver a set power output. We also analyse failure modes of the open-source platform to guide operators toward domain-consistent optima. In application, the domain-enforced platform reduced CO2 emissions by 2.9 kton/y [0.3, 5.4] and lowered annual operating cost by 0.95 million USD [0.3 million USD, 1.6 million USD] for the gas turbine system. We anticipate that the AI-powered simulation platform can adapt to dynamic industrial power-generation environments and broaden access to AI and optimisation tools for data-informed decision-making, enabling robust, transparent, reproducible operator decision support across operations.

The accurate prediction of forming limit diagram (FLD) is crucial in sheet metal forming to prevent localized necking. This study investigates the effects of non-quadratic anisotropy and through-thickness normal (TTN) stress on the prediction of FLD and porosity evolution in ductile porous materials, such as Al-2090 and Al-3104 aluminum alloys. Various non-quadratic anisotropic yield functions (Barlat's yield surfaces YLD-1991, YLD-200418p, YLD-2004-27p, YLD-2011-18p, and YLD-2011-27p) are compared with isotropic and quadratic anisotropic (Hill-1948) yield functions. Gurson-Tvergaard-Needleman model is considered for modeling ductile porous materials. FLDs are predicted using both the bifurcation theory and Marciniak and Kuczynski (M - K) imperfection approach. It is found that Barlat's YLD-1991 shows better agreement with FLD experimental data for Al2090, while Barlat's YLD-2011-27p accurately predicts the FLD for Al-3104. The prediction results also reveal that the presence of TTN stress increases the forming limit strains, especially on the right-hand side of FLD. For the undamaged material model, the FLD predictions show unrealistic sensitivity to TTN stress. By contrast, for the damaged material model, higher TTN stress delays porosity evolution, thus increasing the forming limits. Moreover, it is shown that TTN stress strongly influences void growth, while void nucleation remains unaffected. This study emphasizes the significance of considering non-quadratic anisotropy and damage (porosity), particularly in relation to TTN stress, for accurate FLD prediction in various ductile materials.

A single-source high-flux solar simulator (HFSS) is a crucial device for evaluating solar-sensitive materials and components used in advanced thermal cycles under controlled indoor conditions. This study presents the design, optical analysis, and artificial intelligence-based optimization of a cost-effective 7 kWe HFSS. The optical performance of the system, utilizing both conventional and novel non-ideal segmented reflectors, was evaluated using the Monte Carlo raytracing (MCRT) method and validated against experimental measurements, demonstrating an excellent agreement with only a 1.6% peak flux deviation. The analysis revealed a significant engineering trade-off: the conventional reflector achieves a massive peak flux of 3.419 MW/m2 but suffers from a 55.5% non-uniformity at the focal point, whereas the proposed non-ideal reflector maintains a stable non-uniformity of 16.0–19.6% across various target positions. This non-ideal design offers a robust and resource-efficient alternative without the need for expensive secondary collimators. To overcome the high computational cost of MCRT simulations during setup optimization, a novel predictive High-Flux Solar Simulator Fuzzy Logic (HFSS-FL) model based on Fuzzy Logic (FL) approach is developed in the study. Utilizing a highly resolved Takagi-Sugeno inference system, the FL model instantaneously predicts the peak flux and non-uniformity for any target defocusing distance. The FL model exhibited exceptional accuracy, yielding a ~0% relative error when validated against the extensive MCRT data. Ultimately, this AI-optimized approach enables the rapid determination of optimal testing conditions for high-temperature material stress testing, thereby accelerating the development of high-efficiency solar thermal technologies and contributing to global sustainability and net-zero emission goals.

The domain-consistent and operator-centric artificial intelligence (AI) adoption has remained slow in the industrial operation of power systems, including gas power plants. This paper presents a domain-enforced and operator-in-the-loop neural simulation platform that is built upon embedding the neural surrogate models in the nonlinear optimisation framework and is implemented to analyse the operation of 395 MW capacity gas turbine system. Feed forward architecture-based neural network models like Artificial Neural Network (ANN), Data Information integrated Neural Network (DINN) and Kolmogorov-Arnold Networks (KAN) are trained to predict performance variables of gas turbine system (Power-MW, Turbine Heat Rate-kJ/kWh, Thermal Efficiency-%). KAN achieved slightly higher predictive performance on test dataset (R2 > 0.96) better than those of ANN (R2 > 0.92) and DINN (R2 > 0.93). Mahalanobis distance-based constraint introduces operator-in-the-loop and enforces the data-driven domain for estimating domain-consistent and energy-efficient optimised operating levels to produce a set value of power from gas turbine system. The failure modes of operation of the open-source neural simulation platform are also discussed to guide operators in estimating domain-consistent optimal operating levels. The domain-enforced neural simulation platform can reduce 2.9 kton/y [0.3 kton/y, 5.4 kton/y] of CO2 emissions and may cut the annual operating cost of $ 0.95 m [$ 0.3 m, $ 1.6 m] from the operation of gas turbine system. We anticipate that the developed AI-powered simulation platform may adapt to the dynamic industrial power generation environment and enhance the access of AI and optimisation tools for data-informed decision-making for industrial applications.

Coal-fired power plants emit large quantities of hazardous pollutants including sulfur dioxide (SO₂), oxides of nitrogen (NOx) and Mercury (Hg) that threaten environmental sustainability. Flue gas desulfurization (FGD) systems are widely deployed to reduce SO₂ emissions, yet their performance depends on large number of interacting operational variables, making real-time optimization challenging. This research aims to develop a practical, data-driven optimization framework for performance improvement of industrial-scale FGD systems. Artificial neural network (ANN) based process models have been trained for its proven capability to model complex nonlinear relationships in high-dimensional process data, and reasonable memory requirement for making excellent function approximate for real-life applications. Two years of continuous operational data from a 660 MW coal power plant were used to train ANN models that predict desulfurization efficiency, NOx, and Hg emissions based on key flue gas and slurry parameters. Monte Carlo sensitivity analysis showed that absorber slurry pH, inlet NOx concentration, and inlet dust concentration are the dominant factors for the three outputs, respectively. A Non-Dominated Sorting Genetic Algorithm II (NSGA-II) was applied to determine optimal operating settings under varying plant load scenarios, with the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) selecting the most balanced solutions. Results show that the optimized conditions improve SO₂ removal efficiency while reducing NOx and Hg emissions compared to conventional setpoints. The proposed framework offers a practical pathway for cleaner and more efficient operation of large-scale FGD systems, supporting the power sector’s net-zero objectives.

The rapid evolution of artificial intelligence (AI) and computational methods has transformed the landscape of modeling, simulation, and optimization, particularly for complex, interdisciplinary systems [...]

This research work focuses on binary and novel alloy-target-based ternary nitride low-temperature PVD coatings for implant applications. This study also contributes toward the application of TiAlN/316L coating system for orthopedic implants. TiN/316L, TiAlN(80:20)/316L, TiAlN(70:30)/316L, TiAlN(50:50)/316L coating systems are investigated for bio-tribo-mechanical-corrosive functional properties stack relating patients’ comfort, implants longevity, design configurations, and patient-specific requirements. The Coating systems are characterized by nano-indentation, micro-scratch, pin-on-disk tribometry, XRD, MTT assay, cell adhesion, open circuit potential, electrochemical impedance spectroscopy, and potentiodynamic polarization. Coating systems showed 3.4 to 15.2 times improved surface hardness depending upon their configurations. Overall, maximum coating adhesion of 23.33N is achieved for a TiAlN(70:30)/316L configuration. The bio-corrosive protection efficiency of 91.20, 83.62, 80.92, and 63.99

Rotodynamic analysis is a key analysis for turbomachinery for investigating the health and integrity of equipment. Most of the analyses are performed at the design stage, while the actual machine behavior is different due to imperfections like unbalance, misalignment, cracks, and so forth. In this paper, a representative CAD model of a gas turbine rotor is developed to get the actual rotodynamic response of a rotor. Vibration data of the rotor is compared with that of the developed numerical model. The reference model representation of the actual machine in terms of critical speed and vibration value is found to be 99.7% and 99.1%, respectively. Rotodynamic analyses of numerical models are performed for early identification of faults under various scenarios of unbalance, crack, and crack with unbalance. For these scenarios, modal analysis and harmonic analysis are performed. Natural frequencies and vibration behavior are utilized to capture the variation that indicates the presence of a fault. This way, early identification of faults is made to save the machine from damage. Within the unbalance range of 1.0x10-9 ${1.0\times 10}<^>{-9}$ to 0.5 kg, a direct relation between change in unbalance mass and vibration amplitude is observed in the case of unbalance and unbalance with crack. Similarly, for cracks (of 1-3 mm thickness and depth up to 372 mm), a shift in maximum vibration amplitude frequency to first critical speed from second critical speed is noted. Hilbert transform is utilized to track the nonlinearity especially up to an operating speed of 3000 rpm (50 Hz). These key outcomes can be used to reduce rotary machine downtime by not only highlighting the problem at a very early stage but also swiftly identifying its root cause for the smooth working of rotary equipment in the industry.

Thermal management of photovoltaic systems is important since their electrical output decreases with the increase in temperature. Thermal regulation of a photovoltaic system using a phase change material is an effective technique, but a careful selection of PCM in terms of its thermophysical properties is essential for its better performance. In this work, performance analysis of a novel medium concentrated photovoltaic system employing two mono-facial polycrystalline cells is carried out. The system is thermally regulated with a phase change material. An experimentally validated finite element-based coupled optical, thermal, and electrical model is used to analyze the system's performance. The impact of the thermophysical properties of a PCM such as melting temperature, thermal conductivity, and heat of fusion on the thermal regulation of the system is studied using artificial neural networking methods. The optimum thermophysical properties of the PCM are determined using parametric analysis for the ambient temperatures ranging between 25 and 50 degrees C and a concentration ratio of 20x. Moreover, the performance of the system is analyzed using the optimum PCM for the semi-arid weather conditions of Lahore, Pakistan and the optimum PCM for oceanic weather conditions of Waterford, Ireland. It is found that the melting temperature, thermal conductivity, and heat of fusion of the PCM have a linearly indirect relationship with the temperature of the photovoltaic system while the ambient temperature has a linearly direct relationship with the photovoltaic system's temperature. The melting temperature of a PCM should be 10-15 degrees C higher than the ambient temperatures up to the ambient temperature of 40 degrees C. The melting temperature of a PCM was found to be 5 - 10 degrees C higher than the ambient temperatures for ambient temperatures greater than 40 degrees C. The required thermal conductivity of a PCM increases with the increase in ambient temperature ranging from 10 to 12 Wm(-1)K(-1) for the ambient temperature of 25 degrees C while 18-20 Wm(-1)K(-1) for the ambient temperature of 50 degrees C. The suitable heat of fusion is found in the range of 210-220 kJkg(-1). It is found that the optimum PCM for Lahore has melting temperature, thermal conductivity, and heat of fusion of 53-56 degrees C, 19 Wm(-1)K(-1) and 220 kJkg(-1) respectively. The optimum PCM for Waterford has melting temperature, thermal conductivity, and heat of fusion of 35-37 degrees C, 11 Wm(-1)K(-1) and 220kJkg(-1) respectively. The maximum temperature of concentrated photovoltaic cell for Lahore, Pakistan remains below 83 degrees C, while for Waterford, Ireland, it is below 59 degrees C for all the months in a year. From April to August, the output power is higher for Waterford with an average difference of 12%, while from September to March, Lahore has the higher power output with an average difference of 47%. The maximum power obtained for Lahore is 0.185kWh/day/m(2), while for Ireland it is 0.213kWh/day/m(2). A maximum deviation of 6% is found for electrical output and less than 3% for thermal output between simulated and experimental results during validation. [GRAPHICS] .

The fast depletion of conventional fuel supplies has forced the world to find suitable substitutes to overcome the expected energy crisis. Fossil fuels also contribute to global warming because of their harmful emissions. Biofuels are sustainable and environment friendly. Biodiesel can be sourced from both edible and non-edible oils to replace fossil fuels. To avoid a shortage of food supply, it is preferred to produce biodiesel from non-edible oils. In this research, Litchi chinensis seed oil (LSO) is used as a feedstock to synthesize biodiesel employing transesterification using a microwave oven. The catalyst, potassium hydroxide (KOH), used in this research was extracted from potato waste. Sun-dried potato waste was converted into ash. The produced ash is then dissolved in distilled water, leading to a 34% yield of KOH. The transesterification achieves a 92.9% conversion rate under the conditions: 30% microwave power utilization, a catalyst loading of 15% (W/W), a stirring speed of 700 RPM, and a methanol concentration of 70% (V/V) with an 8-min reaction time. Response surface methodology (RSM), in comparison with artificial neural networks (ANNs), has been utilized for the optimization of biodiesel yield, giving efficient results with errors of 0.003% for RSM and 0.005% for ANN. Consequently, the study reports optimized biodiesel yields of 92.9% (experimental), 93.27% (RSM), and 92.40% (ANN). Physicochemical properties such as kinematic viscosity (4.4 mm2/s) at 40°C, density (875 kg/m3) at 15°C, cetane number (53.2), calorific value (38.8 MJ/kg), flash point (175°C), oxidative stability (6.1 h), and cold flow properties were determined with respect to the ASTM and EN standards. The findings reveal that biofuels primarily support Sustainable Development Goals (SDGs) 7 and 13, with the prime focus on “affordable and clean energy” and “climate action,” respectively.

The Aluminum alloy AA7075 workpiece material is observed under dry finishing turning operation. This work is an investigation reporting promising potential of deep adaptive learning enhanced artificial intelligence process models for L18 (6133) Taguchi orthogonal array experiments and major cost saving potential in machining process optimization. Six different tool inserts are used as categorical parameter along with three continuous operational parameters i.e., depth of cut, feed rate and cutting speed to study the effect of these parameters on workpiece surface roughness and tool life. The data obtained from special L18 (6133) orthogonal array experimental design in dry finishing turning process is used to train AI models. Multi-layer perceptron based artificial neural networks (MLP-ANNs), support vector machines (SVMs) and decision trees are compared for better understanding ability of low resolution experimental design. The AI models can be used with low resolution experimental design to obtain causal relationships between input and output variables. The best performing operational input ranges are identified for output parameters. AI-response surfaces indicate different tool life behavior for alloy based coated tool inserts and non-alloy based coated tool inserts. The AI-Taguchi hybrid modelling and optimization technique helped in achieving 26% of experimental savings (obtaining causal relation with 26% less number of experiments) compared to conventional Taguchi design combined with two screened factors three levels full factorial experimentation.

Desalination is among the most feasible solutions to supply sustainable and clean drinking water in water scarcity areas. In this regard, Multi-Effect Desalination (MED) systems are particularly preferred for harsh feeds (high temperature and salinity) because of their robust mode of operation for water production. However, maintaining the efficient operation of the MED systems is challenging because of the large system design and variables' interdependencies that are sensitive to the distillate production. Therefore, this research leverages the power of machine learning and optimization to estimate the optimal operating conditions for the maximum distillate production from the MED system. In the first step, detailed experimentation is conducted for distillate production against hot water temperature (HWT) varying from 38 to 70 degrees C, and feed water temperature (FWT) is changed from 34 to 42 degrees C. Whereas, the feed flow rate (FFR) is investigated to be varied nearly from 3.6 to 8.7 LPM in the three stages, i.e., FFR-S1, FFR-S2 and FFR-S3. The compiled dataset is used to make the process models of the MED system by three ML-based algorithms, i.e., Artificial Neural Network (ANN), Support Vector Machine (SVM), and Gaussian Process Regression (GPR) under rigorous hyperparameters optimization. GPR exhibited superior predictive performance than those of ANN and SVM on R2 value of 0.99 and RMSE of 0.026 LPM. Monte Carlo technique-based variable significance analysis revealed that the HWT has the highest effect on distillate production with a percentage significance of 95.6 %. Then Genetic Algorithm is used to maximize the distillate production with the GPR model embedded in the optimization problem. The GPR-GA driven maximum distillate production is estimated on HWT = 70 +/- 0.5 degrees C, FWT = 40 +/- 2.5 degrees C, FFR-S1 = 6 +/- 2.6 LPM, FFR-S2 = 7 +/- 1 LPM and FFR-S3 = 7 +/- 1. The ML-GA-based system analysis and optimization of the MED system can boost the distillate production that promotes operation excellence and circular economy from the desalination sector.

In recent years, artificial intelligence has become increasingly popular and is more often used by scientists and entrepreneurs. The rapid development of electronics and computer science is conducive to developing this field of science. Man needs intelligent machines to create and discover new relationships in the world, so AI is beginning to reach various areas of science, such as medicine, economics, management, and the power industry. Artificial intelligence is one of the most exciting directions in the development of computer science, which absorbs a considerable amount of human enthusiasm and the latest achievements in computer technology. This article was dedicated to the practical use of artificial neural networks. The article discusses the development of neural networks in the years 1940–2022, presenting the most important publications from these years and discussing the latest achievements in the use of artificial intelligence. One of the chapters focuses on the use of artificial intelligence in energy processes and systems. The article also discusses the possible directions for the future development of neural networks.

Accurately predicting fuel blends' lower heating values (LHV) is crucial for optimizing a power plant. In this paper, we employ multiple artificial intelligence (AI) and machine learning-based models for predicting the LHV of various fuel blends. Coal of two different ranks and two types of biomass is used in this study. One was the South African imported bituminous coal, and the other was lignite thar coal extracted from the Thar Coal Block-2 mine by Sind Engro Coal Mining Company, Pakistan. Two types of biomass, that is, sugarcane bagasse and rice husk, were obtained locally from a sugar mill and rice mill located in the vicinity of Sahiwal, Punjab. Bituminous coal mixture with other coal types and both types of biomass are used with 10%, 20%, 30%, 40%, and 50% weight fractions, respectively. The calculation and model development procedure resulted in 91 different AI-based models. The best is the Ridge Regressor, a high-level end-to-end approach for fitting the model. The model can predict the LHV of the bituminous coal with lignite and biomass under a vast share of fuel blends.

Abstract Microchannel heat exchangers are heat exchangers with a tube diameter of less than 1 mm. Conventional cooling approaches such as the forced‐air cooling technique fail in high technological compact systems because of the small‐sized surfaces of chips and circuits. In comparison, microchannel heat exchangers are being extensively utilized in compact‐sized devices where a high heat‐transfer medium is required. Moreover, consumers' continued desire for compact products has prompted researchers to study microchannel heat exchangers for their ability to boost the rate of heat transfer that ensures the safety of compact designs. This study presents the evaluation of performance parameters and the manufacturing aspects of microchannel heat exchangers. This study also examines how microchannel heat exchangers are affected by several parameters, including the type of working fluid used, Brownian motion, geometry of the channel, Reynolds number, Nusselt number, Knudsen number, wall resistance of the channel, the effect of gravity, and inlet and outlet arrangement for fluid. Investigating the various geometries for the microchannel indicates that the least pressure drop occurs in square shape cross‐section channels while the highest pressure drop occurs in channels with triangular cross‐sections. Moreover, it has been observed that, with the addition of nanoparticles to the working fluid, the thermal properties of the exchangers as well as the pressure drop increases while at the same time it reduces the boundary layer thickness. In addition, the Reynolds number affects the performance irrespective of the channel geometry. When the fluid is added with nanoparticles, like, Al2O3 and copper oxide (CuO) with different volumetric fractions (φ) of 0%, 0.5%, 1%, 1.5%, and 2%, the performance of the microchannel rises with rising the Reynolds number but conversely when the fluid is used in pure form the performance decreases with the rising value of Reynolds number. In addition, it has been observed that the overall improvement is obtained at φ = 2% and Re = 100 for CuO–water nanofluid. Apart from this, the least heat transfer is recorded at φ = 0.5% and Re = 1.00 for both nanofluids. Moreover, this study concludes that the Nusselt number is independent of the Reynolds number in the regime of laminar flow. It is further evident that as the nanoparticle size reduces, the Nusselt number rises when all the remaining conditions are the same. Apart from this, investigating the inlet/outlet arrangement through finite volume method for D‐, I‐, N‐, S‐, U‐, and V‐type arrangements, it is evident that the V‐type sink has overall greater performance. In terms of gravity, it is observed that it has no effect on the performance of microchannel heat exchangers. This study further illustrates the effects of the fabrication method on the performance of the microchannel heat exchangers.

This research work focus on dry finishing (turning) of AA7075. Newly, emerging alloy target–based ternary PVD–coated tool inserts are compared with non-alloy target–based binary and ternary-coated tool inserts. The coatings include novel alloy target coatings TiAlN (80:20), TiAlN (70:30), and TiAlN (50:50) and non-alloy target coatings TiN and TiCN. One uncoated and five coated tungsten carbide (WC) inserts are analyzed for maximum tool life and minimum surface roughness in dry machining (finish turning) process. Depth of cut, feed rate, and cutting speed are operational parameters optimized against performance parameters (tool life and surface roughness) using a two-stage design of experiment approach. Initial experiments are carried out using L18 (6133) special orthogonal array. General linear model (GLM), linear regression model (LR), and stepwise forward regression model (SFR) are used for analysis. Among alloy target–based PVD–coated tool inserts, TiAlN (50:50) and TiAlN (70:30) performed better in terms of tool life and surface roughness respectively. The TiN-coated tool insert is found as overall superior performing tool insert considering the finished products roughness concerns of aero industry. The optimized solution for TiN is obtained by designing a full factorial experimental configuration. At low feed rate and low cutting speed, high feed rate and low cutting speed, and high feed rate and high cutting speed, approximately 45

Pakistan is a developing country that faces severe energy crises due to the increased use of energy. The purpose of this study is energy transition by designing a strategy for the adoption of renewable energy policies in the entire energy system by using all renewable energy resources to forecast future energy needs and carbon emission mitigation potential. This research study aims to evaluate the renewable energy policies of Pakistan and to analyze the ways to secure energy sources in the future using LEAP. The study established a path for the transformation of the Pakistan energy system by considering the potential of renewable resources, the cost of the energy system, and the primary energy supply. The highest value of energy demand is noted for the 1st scenario, while the lowest emissions are noted for the 16th scenario for each renewable source (WIN16, SOL16, and BIO16). The lowest values of energy demand and emissions (192.1 TWh and 37.7 MMT, respectively) are shown using the green solution compared to other scenarios (hydro, nuclear, BAU), concluding that the green solution is the most suitable scenario. The analysis shows, that from a technological and economic perspective, it is possible to carry out transformation with the necessary steps to effectively achieve a renewable energy system. The findings of this study show that the green scenario in Pakistan which has the lowest operational and externality costs is the best choice for the future.

A large power generation facility is a complex multi-criteria system associated with multivariate couplings, high dependency, and non-linearity among the operating variables which present a major challenge to ensure efficient power production. In this research, an integrated artificial intelligence (AI) and response surface methodology (AI-RSM) framework to achieve the efficient power production operation of a 660 MW coal power plant is presented. Two AI algorithms, i.e., extreme learning machine (ELM) and support vector machine (SVM) are trained comprehensively on the power plant's operational data and are validated as well. Full factorial design of experiments on the three levels of the operating parameters are constructed and simulated from the better performing AI model which is an effective non-linear representation of the complex power plant operation. RSM analysis is carried out under three power generation scenarios to simulate the effective values of the operating variables which are tested on the power plant's operation and a reasonable agreement is found with the experimental observations. The notable improvement in fuel consumption rate, thermal efficiency, and heat rate of the power plant under Half Load, Mid Load, and Full Load capacity of the power plant is achieved by the AI-RSM framework enabled analyses. It is estimated that annual reduction in CO2, CH4 and Hg emissions measuring 210 kg tons per year (kt/y), 23.8 t/y and 2.7 kg/y, respectively can be obtained corresponding to Mid Load operating state of the power plant. The research presents the reliable and robust utilization of AI-RSM framework for simulating the effective operating conditions for the fossil-based power plants' operation with an eventual goal to improve the techno-environmental performance which is expected to contribute to net-zero emissions goal from the energy sector.