Optimization Framework Research Articles

Data-driven inverse design of a multiband second-order phononic topological insulator

Second-order phononic topological insulators (SPTIs) have sparked vast interest in manipulating elastic waves, owing to their unique topological corner states with robustness against geometric perturbations. However, it remains a challenge to develop multiband SPTIs that yield multi-frequency corner states using prevailing forward design approaches via trial and error, and most inverse design approaches substantially rely on time-consuming numerical solvers to evaluate band structures of phononic crystals (PnCs), showing low efficiency particularly when applied to different optimization tasks. In this study, we develop and validate a new inverse design framework, to enable the multiband SPTI by integrating data-driven machine learning (ML) with genetic algorithm (GA). The relationship between shapes of scatterers and frequency bounds of multi-order bandgaps of PnCs is mapped via developing artificial neural networks (ANNs), and a multiband SPTI with multi-frequency topological corner states is cost-effectively designed using the proposed inverse optimization framework. Our results indicate that the data-driven approach can provide a high-efficiency solution for on-demand inverse designs of multiband second-order topological mechanical devices, enabling diverse application prospects including multi-frequency robust amplification and confinement of elastic waves.

Autoinjector optimization through cavitation response and severity minimization

Abrupt acceleration of the syringe of an autoinjector (AI) upon rod-plunger impact may induce undesired severe cavitation events and impose extraneous stresses upon the device, leading to device failure. Cavitation results from a rapid and significant pressure drop in a liquid, leading to the formation and growth of small vapor-filled cavities. Upon collapse, these cavities generate an intense shock wave that may lead to protein aggregation and device container damage and shatter. Since the maximum acceleration of the syringe depends upon the operating conditions of the AI, the severity of cavitation will likewise depend on the operating conditions of the AI. Likewise, injection time and ensuring proper needle displacement before drug release also depend on operating conditions, making optimization of the autoinjector a multiobjective optimization problem.Therefore, in this study, optimization of an autoinjector to limit cavitation severity is pursued via an experimentally validated computational model for cavitation in spring-driven autoinjectors. Our goal is to locate AI design configurations that balance maximizing device performance and patient comfort and minimizing the risks of device damage and severe cavitation upon actuation.Relevant parameters of interest are the drive spring force, air gap height, solution viscosity, friction between the rod and spring, frictional force on the plunger, rates of change of frictional force on the plunger, elasticity of plunger, viscosity of the plunger, and initial displacement between the plunger and the driving rod. The kinematics of the syringe barrel, needle displacement (travel distance) at the start of drug delivery, and injection time are gathered using an experimentally validated autoinjector kinematics model. At the same time, cavitation bubble dynamics are resolved using an experimentally validated cavitation model that takes the temporal displacement of the syringe and temporal air gap pressure as inputs.We use our experimentally validated models to explore the parameter space and understand the driving factors of our desired outcomes. Subsequently, we pose the design problem as a multi-objective optimization problem and develop a deep neural network surrogate model supplemented with iterative learning to speed up optimization. A variance-based sensitivity analysis was performed to determine the sensitivity and influence of design parameters on the outcomes, and the main contributors to the outcomes of interest were isolated. Using a multi-objective optimization framework, we located 300 + successful candidates and evaluated them through uncertainty analysis to identify three promising candidates that meet all criteria for drug viscosities of interest. Finally, we show that this methodology can be used to conduct hypothesis testing, leading to novel design configurations.

Surrogate-assisted multi-objective Bayesian optimization for improved rheological design of bioinks

Extrusion bioprinting is a popular biofabrication technique for creating viable biological constructs with applications ranging from regenerative medicine and cancer research to therapeutics screening. A primary challenge within extrusion bioprinting lies in preserving cell viability and function amidst the shear forces generated during extrusion. Therefore, the formulation of shear-thinning bioinks with tailored flow properties is essential to minimize shear stress accumulation during printing. However, this task is complex, as the bioink’s flow properties are contingent upon numerous factors, including but not limited to the polymer concentration, spatial cell density in the bioink, molecular weight of the polymer, and the nozzle’s geometry. The current methods for adjusting flow properties are heuristic, posing challenges to the rapid development models and automation of the bioprinting workflow. In this study, we introduce a Gaussian Process-based multi-objective Bayesian optimization framework that rapidly identifies Pareto-optimal flow property values to minimize hydrodynamic shear stresses during extrusion. The optimization focuses on four flow parameters, minimum viscosity, maximum viscosity, consistency index, and flow behavior index from the power-law model of shear-thinning fluids alongside the nozzle outlet diameter. Following an initial set of 24 experimental runs, our sequential design approach required 21 iterations to achieve convergence on a set of flow parameters that meet the Pareto optimality criteria. The presented strategy, which integrates physics-based numerical models with data-driven optimization models, emerges as a tool for the rapid fine-tuning of bioink flow behaviors. The multi-objective framework also sets the stage for future exploration into cell viability and functional outcomes post-bioprinting.

An Antenna Optimization Framework Based on Deep Reinforcement Learning

An Antenna Optimization Framework Based on Deep Reinforcement Learning

The Nexus of Fiscal Policy and Growth in the Optimal Control Framework

This research investigates the effect of fiscal policy on economic growth, focusing on the intertwined relationship between public goods provision, taxation policies, and their impact on economic dynamics and growth. The study aims to contribute to this research domain by introducing an economic model within the optimal control framework that integrates taxation policies into the social constraint and incorporates public goods into the social utility function, addressing recent limitations in this research area. Through numerical analysis, the study examines the potential effects of taxes and public goods provision on consumption and economic growth, as reflected by the level of capital. Results indicate that increasing government spending on public goods can decrease private goods consumption without significant economic growth benefits. At the same time, high tax rates could potentially hinder economic growth by overly relying on government intervention. The research findings highlight the importance of an optimal approach to fiscal policy, with policy implications including the need to carefully evaluate public funds' allocation, enhance public spending efficiency, and implement optimal tax to foster economic growth.

Assessing the Recreational Fishers and their Catches based on Social Media Platforms: Privacy and Ethical Data Analysis Considerations

Social Media (SM) can offer insights to track recreational Fishing (Rf); Many Limitations Hinder Its Practical Application. The proportion and characteristics of RFs that share their catches remain unidentified. The objective was to enhance the surveillance abilities of RF by utilizing SM information. This study focuses on the sharing economy information transmission, network optimization of SM platforms, and privacy protection issues during data transmission. The study starts with the data transmission characteristics in SM platforms, analyzes the data ethics in SM information transmission from the perspectives of natural and economic data attributes, and then focuses on the privacy protection principles in network information transmission. Then, a privacy protection scheme based on a locally sensitive hash algorithm is constructed, and finally, an information transmission scheme and method optimization based on a network optimization module are proposed. The research gathered data using physical (face-to-face) surveys and digital (email) surveys to define marine RF that post catches on online platforms (“sharers”), together with additional demographic and fishing data. A comparative analysis was conducted on the computational convergence and accuracy of different privacy protection methods, and the optimal rule framework for data privacy protection was discussed. The observation results show that the efficiency of information transmission using the Minhash-SSNR method is slightly inferior to that using the OSNR-SSNR method. Minhash technology is based on a similarity comparison of dimensionality-reduced data. This leads to data causing the initially highly similar two sets of lists to be misjudged as not having enough similarity, thereby reducing some potential information dissemination opportunities and affecting the efficacy of information transmission. It can be seen that the Minhash-SSNR strategy can effectively send information to nodes with high similarity, preventing excessive information duplication within the system. Although the Minhash-SSNR strategy has a certain degree of decline in information transmission efficiency, it accounts for only about 5% compared to the OSNR-SSNR strategy, ensuring the essential operational stability of SM opportunity networks without significant impact. With few learning and training times, the network information transmission optimization module proposed in this study quickly achieved a lower exponential error. As the number of training sessions gradually increases, the exponential error of the network information transmission optimization module in this study is small, and the prediction accuracy of the method is high. RFs who captured a prize, iconic, or symbolic species tended to discuss their catches more. This research signifies significant progress in incorporating SM information to oversee RFs.

A seismic-risk-based bi-objective stochastic optimization framework for the pre-disaster allocation of earthquake search and rescue units

Accurately predicting earthquakes' time, location and size is nearly impossible with today’s technology. Severe earthquakes require prompt and effective mobilization of available resources, as the speed of intervention has a direct impact on the number of people rescued alive. This, in turn, calls for a strategic pre-disaster allocation of search and rescue (SAR) units, both teams and equipment, to make the deployment of resources as quick and equitable as possible. In this paper, a seismic risk-based framework is introduced that takes into account distance-based contingencies between cities. This framework is then integrated into a mixed-integer non-linear programming (MINLP) problem for the allocation of SAR units under uncertainty. The two minimization objectives considered are the expected maximum deployment time of different SAR units and the expected mean absolute deviation of the fulfillment rates. We recover the best vulnerability-adjusted routes for each size-location scenario as input to the optimization model using the dynamic programming (DP) approach as part of the broader area of reinforcement learning (RL). The results of the hypothetical example indicate that the comprehensive model is feasible in various risk scenarios and can be used to make allocation-deployment decisions under uncertainty. The results of the sensitivity analysis verify that the model behaves reasonably against changes in selected parameters, namely the number of allowed facilities and weights of individual objectives. Under the assumption that the two objectives are equally important, the model achieves a total deviation of %3.5 from the objectives with an expected maximum dispatch time of 1.1327 hours and an expected mean absolute deviation of 0.01.

Fruit and vegetable leaf disease recognition based on a novel custom convolutional neural network and shallow classifier.

Fruits and vegetables are among the most nutrient-dense cash crops worldwide. Diagnosing diseases in fruits and vegetables is a key challenge in maintaining agricultural products. Due to the similarity in disease colour, texture, and shape, it is difficult to recognize manually. Also, this process is time-consuming and requires an expert person. We proposed a novel deep learning and optimization framework for apple and cucumber leaf disease classification to consider the above challenges. In the proposed framework, a hybrid contrast enhancement technique is proposed based on the Bi-LSTM and Haze reduction to highlight the diseased part in the image. After that, two custom models named Bottleneck Residual with Self-Attention (BRwSA) and Inverted Bottleneck Residual with Self-Attention (IBRwSA) are proposed and trained on the selected datasets. After the training, testing images are employed, and deep features are extracted from the self-attention layer. Deep extracted features are fused using a concatenation approach that is further optimized in the next step using an improved human learning optimization algorithm. The purpose of this algorithm was to improve the classification accuracy and reduce the testing time. The selected features are finally classified using a shallow wide neural network (SWNN) classifier. In addition to that, both trained models are interpreted using an explainable AI technique such as LIME. Based on this approach, it is easy to interpret the inside strength of both models for apple and cucumber leaf disease classification and identification. A detailed experimental process was conducted on both datasets, Apple and Cucumber. On both datasets, the proposed framework obtained an accuracy of 94.8% and 94.9%, respectively. A comparison was also conducted using a few state-of-the-art techniques, and the proposed framework showed improved performance.

An optimized deep hybrid learning for multi-channel EEG-based driver drowsiness detection

Driver drowsiness is a severe issue that has contributed to numerous fatal accidents and injuries. Thus, detecting driver drowsiness is an important task that has been the subject of intensive research in recent years. There have been numerous proposed physiological signals for detecting driver drowsiness. However, the Electroencephalographic (EEG) signal is the most commonly used due to its direct relationship with drowsiness and its simplicity of acquisition. Recently, different Machine Learning (ML) and Deep Learning (DL) models have been proposed to detect driver drowsiness. This study utilized a publicly accessible dataset containing twelve healthy participants. Reading numerous research papers, we determined no specific EEG-based drowsiness preprocessing parameter values. Consequently, as a first step, and for the first time, to our knowledge in this field, we applied an optimization algorithm to determine the optimal preprocessing parameter values using a CNN model and accuracy as the objective function. The obtained results demonstrated the importance of selecting the correct values, as the mean accuracy score increased from 91% before optimization to 95% after optimization for the proposed CNN. The training time has been reduced. Also, as a second step, we have used the Optuna Hyperparameter optimization framework to select the optimal CNN Hyperparameters, which increased the mean accuracy from 95% to 97%. Finally, to take advantage of Deep Hybrid learning, we have fused the CNN “features extractor” with seven ML classifiers, with the CNN-SVM classifier achieving the highest average accuracy of 99.9%, and the training time has been reduced to a shallow value.

Reverberation room design optimisation for low-frequency diffuse sound absorption testing

The reproducibility of sound absorption testing with the reverberation room method is a long-standing concern. Absorptive samples induce directionality in the nearfield, while the farfield depends on the room geometry below the Schroeder frequency. Nevertheless, when properly accounting for nearfield effects, the theoretical diffuse absorption coefficient of a sample still represents its average performance across an ensemble of different rooms, even at very low frequencies. Recent research found that particular reverberation room designs allow for an accurate measurement of the diffuse sound absorption coefficient of highly absorptive samples at low frequencies. Pinpointing such designs hence opens up a possibility to sustainably improve the low-frequency reproducibility of sound absorption testing in reverberation rooms. The present paper introduces a numerical optimisation framework that serves this purpose. Specific room shapes are parametrised and the geometrical room parameters are optimised so as to minimise the difference between the measured and the diffuse absorption coefficient under appropriate constraints. The sound absorption testing of a sample in a particular reverberation room is numerically simulated using a method that is both accurate and computationally efficient at low frequencies. The diffuse absorption is computed with a hybrid deterministic-statistical energy analysis approach that accounts for the detailed absorber properties, geometry, and boundary conditions, as well as the nearfield effects. The methodology is applied to both cuboidal and hexahedral room shapes. Certain optimised designs are found not only to provide an excellent match for the absorber that was used during the optimisation, but they also maintain their performance across a range of absorptive samples. Additionally, potential geometrical deviations are found to be well tolerated by these reverberation room designs.

Coupling geospatial suitability simulation and life cycle carbon emissions towards uncertain optimization planning for wind-photovoltaic-hydrogen multi-energy complementary system

Building clean and modern multi-energy complementary systems is a direct path to decarbonization in energy sector. The introduction of hydrogen energy with high operational flexibility enables the multi-energy complementary system to fully accommodate renewable energy. The key issues of the optimization planning of the wind-photovoltaic-hydrogen multi-energy complementary system that need to be solved are how to finely characterize the carbon reduction potential of the system, consider the uncertainty of supply and demand in the modeling process, and incorporate the energy supply business of electric vehicles into the planning. This study constructs optimization planning model with the following three progressive stages: geospatial suitability simulation based on geographic information system, uncertainty analysis based on probabilistic scene generation technology, and capacity configuration optimization. Unlike previous studies, this model measures and optimizes the carbon emissions level of the system from a lifecycle perspective. The empirical study results indicate that carbon emissions optimization from the perspective of the entire life cycle would directly affect the system configuration results. From a lifecycle perspective, the carbon emissions accounting result is about 9 times higher than that obtained only during the operation phase. Compared to the traditional distribution system, the carbon emissions of the system constructed in this study decreased by 82.75 %. In addition, considering that the system provides charging services for electric vehicles, which can reduce carbon emissions by about 93.6 % but only increase costs by about 29.4 %. The computational results justify the proposed optimization framework. This study provides optimization models for the system, and the implementation framework can support technical reference for practice.

Flexibility-Oriented AC/DC Hybrid Grid Optimization Using Distributionally Robust Chance-Constrained Method

With the increasing integration of stochastic sources and loads, ensuring the flexibility of AC/DC hybrid distribution networks has become a pressing challenge. This paper aims to enhance the operational flexibility of AC/DC hybrid distribution networks by proposing a flexibility-oriented optimization framework that addresses the growing uncertainties. Notably, a comprehensive evaluation method for operational flexibility assessment is first established. Based on this, this paper further proposes a flexibility-oriented operation optimization model using the distributionally robust chance-constrained (DRCC) method. A customized solution method utilizing second-order cone relaxation and sample average approximation (SAA) is also introduced. The results of case studies indicate that the flexibility of AC/DC hybrid distribution networks is enhanced through sharing energy storage among multiple feeders, adaptive reactive power regulation using soft open points (SOPs) and static var compensators (SVCs), and power transfer between feeders via SOPs.

Parametric Optimization for Fully Fuzzy Linear Programming Problems with Triangular Fuzzy Numbers

This paper presents a new approach for solving FFLP problems using a double parametric form (DPF), which is critical in decision-making scenarios characterized by uncertainty and imprecision. Traditional linear programming methods often fall short in handling the inherent vagueness in real-world problems. To address this gap, an innovative method has been proposed which incorporates fuzzy logic to model the uncertain parameters as TFNs, allowing for a more realistic and flexible representation of the problem space. The proposed method stands out due to its integration of fuzzy arithmetic into the optimization process, enabling the handling of fuzzy constraints and objectives directly. Unlike conventional techniques that rely on crisp approximations or the defuzzification process, the proposed approach maintains the fuzziness throughout the computation, ensuring that the solutions retain their fuzzy characteristics and better reflect the uncertainties present in the input data. In summary, the proposed method has the ability to directly incorporate fuzzy parameters into the optimization framework, providing a more comprehensive solution to FFLP problems. The main findings of this study underscore the method’s effectiveness and its potential for broader application in various fields where decision-making under uncertainty is crucial.

A study of powertrain parameter optimization based on performance requirements of a vehicle platform

Powertrain top down design is a vital process for vehicle platformization and performance development. To obtain the optimal component size of a plug-in hybrid electric vehicle (PHEV) platform, this study proposes an optimization framework with the goal of minimizing performance redundancy. First, we analyze the characteristics of the powertrain architecture, and propose the indicators through performance gradient planning of the vehicle platform. Then we establish the mathematical model of the powertrain, and formulate the optimization problem. Particle Swarm Optimization (PSO) is used to solve the problem and a platform optimization problem with eight vehicle models is performed. Optimization results show that compared to traditional single vehicle powertrain optimization, platform based powertrain optimization can greatly reduce the number of components and save development costs through the sharing of components. With only three electric machines, two engines, two generators, and two sets of transmissions, the performance requirements of the entire platform can be met, and the maximum performance redundancy of different vehicle models is only 0.205. It can be seen that platform based powertrain optimization can greatly reduce vehicle costs and improve profit margins.

Advancing ESG Portfolio Optimization: Methods, Progress, and Future Directions

Objective - The integration of environmental, social, and governance (ESG) criteria into investment portfolios has emerged as a critical field of study, underscoring the interconnectedness between financial markets and global sustainability objectives. Methodology/Technique - This systematic literature review analyzes 157 academic documents, focusing on ESG portfolio optimization methodologies and identifying emerging trends. Key methods reviewed include genetic algorithms, dynamic optimization models, multi-objective optimization frameworks, and machine learning techniques. Findings - Despite considerable advancements, gaps remain, such as the need for broader application across diverse markets and asset classes, improved risk-return assessments, and standardized ESG data reporting. Future research should also investigate the role of central banks and regulators in fostering sustainable finance. Novelty - By addressing these gaps, stakeholders can better align investment practices with sustainability goals, contributing to a more resilient and inclusive global economy. Type of Paper: Review JEL Classification: G11, Q56, G28, G32 Keywords: Sustainable Investment, Sustainable Finance, ESG Portfolio Performance, ESG Risk Management, ESG Portfolio Optimization Reference to this paper should be referred to as follows: Billah, A.L; Koesrindartoto, D.P; Faturohman, T. (2024). Advancing ESG Portfolio Optimization: Methods, Progress, and Future Directions, Acc. Fin. Review, 9(2), 65 – 73. https://doi.org/10.35609/afr.2024.9.2(2)

Which classifiers are connected to others? An optimal connection framework for multi-layer ensemble systems

Ensemble learning is a powerful machine learning strategy that combines multiple models e.g. classifiers to improve predictions beyond what any single model can achieve. Until recently, traditional ensemble methods typically use only one layer of models which limits the exploration of different aspects in the classifiers’ predictions. On the other hand, the rise of deep learning has introduced multi-layer architectures that can learn complex functions by transforming data into multiple levels of representation. This characteristic of deep learning suggests that multi-layer ensembles may potentially provide better performance compared to single-layer ensembles. However, a problem which might arise is that in the subsequent layers, not all the inputs to a classifier are desirable, leading to lower performance. In this paper, we introduce a novel multi-layer ensemble of classifiers named COME in which each classifier at a specific layer is connected to multiple classifiers in the previous layer. These connections signify the use of the previous-layer-classifiers’ outputs as inputs for training the current layer's classifier. Each classifier can be connected to different classifiers in the previous layer, which allows inputs in each layer to be optimally selected. We propose a binary encoding scheme to encode the topology of the proposed multi-layer ensemble with defined connections between layers. Differential Evolution, a popular evolutionary computation method, is used as the optimisation algorithm to search for the optimal set of connections. Experimental results on 30 datasets from the UCI Machine Learning Repository and OpenML demonstrate that our proposed ensemble outperforms many state-of-the-art ensemble learning algorithms.

Ai-enhanced thermal modeling for integrated process-product-system optimization in zero-defect manufacturing chains

This paper proposes a novel approach for realising zero-defect manufacturing by integrating AI-based thermal modelling to perform multi-stage process, product, and system optimisation in complex manufacturing chains. It integrates advanced thermal simulation and sensor data feed in real-time as a multistage thermal prediction and management system via machine learning algorithms. This framework seeks to predict thermal behaviour using a combination of convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and graph neural networks (GNNs) for encoding, predicting, and learning the spatial, temporal, and interactive thermal behaviour, respectively. It further embeds finite element analysis (FEA) simulations for high-fidelity thermal predictions using data fusion through Kalman filters. This helps obtain the optimal estimates of thermal states from sensor measurements involving different types of sensors and the characteristics of signals. A multistage multimodal optimization framework involves genetic algorithms (GA) for global thermal parameter optimisation, reinforcement learning (RL) for multi-stage dynamic process control optimisation, and multi-agent systems (MAS) for coordinated multi-stage multi-objective balance embedded in a digital twin architecture. Evaluation results show that the effectiveness of the proposed system in improving the overall production efficiency is 33%, reducing defects is 47%, and reducing energy utilisation is 22%, when compared to the current de facto approaches. There is also a 38% improvement in predictive capability in preventing, detecting, and predicting of cross-stage process faults.

Energy system optimization based on fuzzy decision support system and unstructured data

To address the complex challenges in energy systems, this study proposes a novel optimization framework that integrates fuzzy decision support and unstructured data processing technologies. This framework aims to improve efficiency, reduce costs, decrease environmental impact, increase system flexibility, and enhance user satisfaction, thereby promoting sustainable development in the energy industry. The framework combines the innovative Energy Semantic Mapping Model (ESMM) and the advanced deep learning architecture ResNet to process textual and visual data effectively. ESMM enables accurate prediction of energy demand, while ResNet significantly reduces equipment maintenance costs and improves energy distribution efficiency. These advancements are critical as they address the limitations of existing approaches in handling large-scale unstructured data and making informed decisions under uncertainty. The Environmental Impact Assessment (EIA) confirms the model's effectiveness in reducing carbon emissions. A comprehensive economic analysis demonstrates substantial cost savings in energy procurement and operations and maintenance, with overall savings exceeding 25%. Enhanced user satisfaction and reduced system response times further validate the practical utility of the proposed approach. Additionally, a genetic algorithm is used to optimize the fuzzy rule base, enhancing the robustness and adaptability of the model. Experimental results show superior performance compared to traditional systems, providing strong empirical evidence for the intelligent transformation of energy systems. This research contributes to the field by offering a more sophisticated and flexible solution for managing energy systems, particularly in terms of leveraging unstructured data and improving decision-making processes.

Adaptive robust optimization framework for market-based wind power investment

Wind is distinguished by its eco-friendliness and sustainability, making it one of the most rapidly expanding forms of renewable energy sources (RESs). Hence, it is necessary to determine the most profitable plan for wind farm installation. This paper constructs a novel scheme for market-based wind power investment (WPI) problems using adaptive robust optimization (ARO). A tri-level robust WPI (RWPI) model is established, the first level of which is to minimize the investment cost plus the worst-case loss. In the second level, the worst-case loss (also known as the maximum regret) is identified by maximizing the minimum value of minus profit over the uncertainty sets. The third level maximizes the wind farm profit. Since the profit calculation requires the determination of the locational marginal price (LMP), the third level constitutes bi-level programming, with the upper level being the profit maximization and the lower level being the market clearing process. First, Karush-Kuhn-Tucker (KKT) conditions are applied to convert the bi-level model to a single-level model, resulting in an ARO with binary variables at the third level. Afterward, the nested column-and-constraint generation (NCCG) strategy is employed to solve the ARO with mixed-integer recourse. A case study is used to verify the scalability and practical applicability of the proposed model.

Integrated batch production planning and scheduling with machine-learning based production capacity region considering wind energy system

Continued greenhouse gas emission has led to increased global warming. Towards the goal of carbon emissions reduction and environmental sustainability, replacing high-polluting fossil fuel by clean energy has become a top priority in the decision-making of the current chemical industry. This work presents an optimization framework to address the integration of batch chemical production planning and scheduling considering energy consumption, which is a vital operational problem in chemical production. Assisting the linkage between planning layer and scheduling layer, a novel linear support vector machine based production capacity region calculation method is proposed to identify feasible regions of the scheduling layer. Uncertain factors in both the production system and the energy system are considered, including demand volatility of chemical products, energy supply fluctuations and time-varying energy market prices. To evaluate each strategy and the proposed model, simulation experiments are performed in three representative case studies. The results reveal that the integrated model considering energy consumption shows superior performance in different energy configurations, with or without wind energy generation and battery storage systems.