Role In Performance Optimization Research Articles

Adaptive modification of turbofan engine nonlinear model based on LSTM neural networks and hybrid optimization method

An accurate and reliable turbofan engine model which can describe its dynamic behavior within the full flight envelop and lifecycle plays a critical role in performance optimization, controller design and fault diagnosis. However, due to the performance differences caused by the tolerance of engine manufacturing and assembly, and performance degradation during continuously stringent environmental regulations, the model accuracy is severely reduced. In this paper, an adaptive modification method of turbofan engine nonlinear Component-Llevel Model (CLM) based on Long Short-Term Memory (LSTM) Neural Network (NN) and hybrid optimization algorithm is pro-posed. First, a dynamic compensator with a combined LSTM NN architecture is constructed to compensate for the initial error between the experimental data and CLM of a turbofan engine under health condition. Then, a sensitivity analysis approach based on the entropy coefficient and technique for order preference by similarity to an ideal solution integrated evaluation is developed to choose the unmeasurable health parameters to be adjusted. Finally, a parallel hybrid optimization algorithm is developed to complete the adaptive model modification when the performance degrades. The proposed method is verified on a military low-bypass twin-spool turbofan engine, and the experimental results show the effectiveness of the proposed method.

The Computer Intelligent Selection of Scientific Research Subjects Through Ensemble Learning for Large-Scale Data Sources and Deep Neural Network

Selecting a proper scientific research subject is critical for scientific researchers and managers. Scientific researching data are from massive sources and have various attributes. For the problem of subject selection, feature extraction and prediction model play important role in performance optimization. In the paper we introduce ensemble learning method to help find the best fit attributes describing data. Our ensemble learning models include random forests, support vector machine, Boltzmann machine and decision tree. Since the data are from many data sources, we adopt multiple models of deep neural network. An acceleration method is used to reduce the training time as well. Experiments shows that the proposed approach performs better than RNN algorithm both in accuracy ratio and recall ratio. The model selection module and acceleration method help optimize the time cost largely.

Core Temperature Response During the Marathon Portion of the Ironman World Championship (Kona-Hawaii).

The Ironman triathlon consists of a 3.8 km swim, 180 km bike, and 42.195 km run. Thermoregulation responses play an important role in performance optimization and injury prevention. Factors such as environmental conditions including heat and humidity, athlete training level, and race duration can affect thermoregulation. Hyperthermia occurs when the core temperature rises above 38.5°C. The present study aims to describe core temperature (Tcore) in top-level and well-trained age group triathletes during the marathon of Ironman World Championship 2014 in Kona-Hawaii under thermal stress conditions. Tcore of 15 triathletes (age: 36.11 ± 7.36 years, body mass: 71.14 ± 7.12 kg, height: 179 ± 0.04 cm, and fat %: 8.48 ± 0.85) who classified for the Ironman World Championship was measured by an ingestible pill telemetry system prior to competition, during the marathon and 60 min after finishing the race. Mean wet bulb globe temperature (WBGT) during the marathon was 24.66°C (range 22.44–28.50°C). Body mass index (BMI) and perceived exertion (Borg Scale and Visual Analog Scale-Pain) were collected before the race and 60 min after the event. Time variables were extracted from their official race time and split times. Finish time was 10: 06:56 ± 0:48:30. Tcore was initially 36.62 ± 0.17°C, increased at the end of the event (38.55 ± 0.64; p < 0.01) and remained elevated 60 min after the event (38.65 ± 0.41°C; p < 0.002). BMI significantly decreased after the event (22.85 ± 1.11 vs. 21.73 ± 1.36; p < 0.05), whereas both exercise perceived exertion [Borg Scale (10.2 ± 1.64 vs. 18.60 ± 1.67; p < 0.003)] and perceived muscle pain [VAS Pain (2.75 ± 1.59 vs. 9.08 ± 1.13; p < 0.001)] increased significantly after the event. Tcore during competition correlated negatively with position in age group (r − 0.949, p = 0.051), but not with race time (r = −0.817; p = 0.183). High-level age group triathletes competing under thermal stress conditions in the Kona Ironman reached a state of hyperthermia during the marathon. After 60 min of recovery the hyperthermia persisted. Strategies to aid post-event cooling and recovery should be considered to avoid the potentially dangerous adverse health effects of hyperthermia.

An adaptive equivalent consumption minimization strategy for plug-in hybrid electric vehicles based on traffic information

Energy management strategies play an important role in performance optimization of plug-in hybrid electric vehicles (PHEVs), and can be further improved by incorporating external traffic information. Motivated by this, an adaptive equivalent consumption minimization strategy considering traffic information is proposed in this study to facilitate the effective energy management of PHEVs. First, the initial equivalent factors in terms of different initial state of charge (SOC) and driving distance are searched by genetic algorithm. Then, the simplified dynamic programming is leveraged to determine the optimal SOC trajectory according to the traffic information with fast calculation speed. The fuzzy controller is employed to regulate the equivalent factor dynamically, thus enabling effective tracking of the reference SOC trajectory. A hardware-in-the-loop simulation platform based on the virtual scene is developed to validate the performance of controller. Simulation and experimental results highlight that the proposed strategy can lead to less fuel consumption, compared to traditional equivalent consumption minimization strategy, thereby proving its feasibility.

Trip-Oriented Model Predictive Energy Management Strategy for Plug-in Hybrid Electric Vehicles

Energy management strategies play a critical role in performance optimization of plug-in hybrid electric vehicles (PHEVs). In order to attain effective energy distribution of PHEVs, a predictive energy management strategy is proposed in this study based on real-time traffic information. First, an exponentially varied model for the velocity prediction is established, of which the tunable decay coefficient is regulated by the supported vector machine (SVM). In this manner, the prediction precision is improved. Then, by properly simplifying the powertrain model, the state of charge (SOC) reference trajectory is generated based on the fast dynamic programming (DP) with fast calculation speed and consideration of the traffic information. Moreover, the typical DP algorithm is leveraged to solve the nonlinear rolling optimization problem for minimizing the operating cost in a receding horizon. Simulation results demonstrate that the proposed algorithm can reach 92.83% operating savings, compared with that of the traditional DP; and save 6.18% cost compared with the MPC algorithm only with reference of the trip duration.

The Yttrium Effect on Nanoscale Structure, Mechanical Properties, and High-Temperature Oxidation Resistance of (Ti0.6Al0.4)1–x Y x N Multilayer Coatings

As machine tool coating specifications become increasingly stringent, the fabrication of protective titanium aluminum nitride (Ti-Al-N) films by physical vapor deposition (PVD) is progressively more demanding. Nanostructural modification through the incorporation of metal dopants can enhance coating mechanical properties. However, dopant selection and their near-atomic-scale role in performance optimization is limited. Here, yttrium was alloyed in multilayered Ti-Al-N films to tune microstructures, microchemistries, and properties, including mechanical characteristics, adhesion, wear resistance, and resilience to oxidation. By regulating processing parameters, the multilayer period (Λ) and Y content could be adjusted, which, in turn, permitted tailoring of grain nucleation and secondary phase formation. With the composition fixed at x = 0.024 in (Ti0.6Al0.4)1–x Y x N and Λ increased from 5.5 to 24 nm, the microstructure transformed from acicular grains with 〈111〉 preferred orientation to equiaxed grains with 〈200〉 texture, while the hardness (40.8 ± 2.8 GPa to 29.7 ± 4.9 GPa) and Young’s modulus (490 ± 47 GPa to 424 ± 50 GPa) concomitantly deteriorated. Alternately, when Λ = 5.5 nm and x in (Ti0.6Al0.4)1–x Y x N was raised from 0 to 0.024, the hardness was enhanced (28.7 ± 7.3 GPa to 40.8 ± 2.8 GPa) while adhesion and wear resistance were not compromised. The Ti-Al-N adopted a rock-salt type structure with Y displacing either Ti or Al and stabilizing a secondary wurtzite phase. Moreover, Y effectively retarded coating oxidation at 1073 K (800 °C) in air by inhibiting grain boundary oxygen diffusion.

Performance evaluation for on-demand routing protocols based on OPNET modules in wireless mesh networks

In wireless networks, users expect to get access to the network securely and seamlessly to share the data flow of access points anytime and anywhere. However, either point-to-point or point-to-multipoint methods in traditional wireless networks make the network bandwidth decrease rapidly, which cannot meet the requirements of users. Recently, a new wireless broadband access network, wireless mesh networks (WMNs), has emerged. As one of the key technologies in WMNs, wireless routing protocols plays an important role in performance optimization of WMNs. Therefore, in this paper, we address the on-demand routing protocols by focusing on dynamic source routing (DSR) protocol and ad hoc on-demand distance vector (AODV) routing protocol in WMNs. Then, we use the OPNET modules to establish the simulation models of DSR and AODV protocols in WMNs. Simulation and results show that, DSR protocol that is based on the dynamic source routing is not suitable for wireless transmission, while AODV routing protocol that is based on the purpose-driven routing is suitable for wireless transmission with rapid change of network topology.

Endoreversible terrestrial solar energy conversion

The analysis of reversible radiation conversion is insightful in process optimization and provides the upper limit to performance of any conceptual device. However, in practice, irreversibilities are unavoidable and play an important role in performance optimization. The finite absorption of a radiative source flux, and the simultaneous emission of radiation, is an inherently irreversible process. Likewise, heat rejection from the conversion device is unavoidable and dependent on local environmental conditions and resources. The endoreversible model treats these irreversibilities as inherent but external to the conversion process. In this paper, the effect of these irreversibilities on performance is investigated for a model with irreversible radiative absorption combined with irreversible conductive or convective heat rejection. Previous models either do not include the magnitude of entropy rejection required by the second law or do not otherwise accurately represent terrestrial solar conversion. Analysis of the model provides a guide for optimal operating conditions. For relatively poor heat rejection, as can occur in arid climates, analysis of the model reveals regions of operating parameters that will result in zero or low theoretical work output, and/or high operating temperatures and increased risk of failure. This is of particular concern for photovoltaic systems with low maximum temperature limits. Analytical expressions for maximum ideal work are provided, given the specific radiative source flux and heat rejection conditions. A fair evaluation of system performance is thereby obtained by comparing actual work production to the ideal, given the specific external operating conditions or restrictions.

A unified approach to Markov decision problems and performance sensitivity analysis with discounted and average criteria: multichain cases

We propose a unified framework to Markov decision problems and performance sensitivity analysis for multichain Markov processes with both discounted and average-cost performance criteria. With the fundamental concept of performance potentials, we derive both performance-gradient and performance-difference formulas, which play the central role in performance optimization. The standard policy iteration algorithms for both discounted- and average-reward MDPs can be established using the performance-difference formulas in a simple and intuitive way; and the performance-gradient formulas together with stochastic approximation may lead to new optimization schemes. This sensitivity-based point of view of performance optimization provides some insights that link perturbation analysis, Markov decision processes, and reinforcement learning together. The research is an extension of the previous work on ergodic Markov chains (Cao, Automatica 36 (2000) 771).