On-board real-time optimization control for turbofan engine thrust under flight emergency condition

Fang

et al. 2019

IEEE Access

Self Cite

A new on-board turbo-fan engine modeling method based on a batch normalize (BN) minibatch gradient descent (MGD) deep neural network (NN) is proposed. This new method adopts BN algorithm, which accelerates the network training speed and overcomes the gradient vanish problem. Hence, using the BN algorithm, the neural network adopts the deeper structure, which means the network has a stronger representation capacity. This mini-batch gradient descent (MGD-NN) algorithm that consumes much less time to update the NN parameters is adopted. Therefore, it is more suitable for training big dataset and establishing a high-accuracy engine model in a large flight envelope. Finally, to verify whether the proposed method could be applied to larger flight envelope, the conventional NN also adopts MGD (called MGD-NN). The turbo-fan engine models based on these two modeling methods are both conducted within a sub-sonic cruise envelope. The simulation results show that the proposed modeling method has much higher accuracy than the MGD-NN. Moreover, the proposed method has the characteristics of less data storage, low computation complexity, and good real-time performance, which are the most importance indices for model realize on-board. INDEX TERMS Aero-engine model, batch normalize, deep neural network, turbo-fan on-board model, mini-batch gradient descent, data storage.

“…The principle of back-propagation is shown in Figure 4. Supposingδ n net as: (15) where n net is the number of network layer, ⊗ is hadamard product, that is Supposing δ l :…”

Section: Back-propagation Algorithmmentioning

confidence: 99%

Aero-Engine On-Board Model Based on Batch Normalize Deep Neural Network

Fang

et al. 2019

IEEE Access

Self Cite

“…The simulation object adopted is a bench mark model based on CLM for a turbo-fan engine which has high fidelity in dynamic and static performance related to its real one. 22 The simulation processes of these two methods choose acceleration process. The starting point and ending point of acceleration are the steady operating point when power level angle PLA is 26° and 70°, respectively.…”

Section: The Simulations Of Nmpcmentioning

confidence: 99%

“…The engine nonlinear model based CLM estimates the unmeasurable parameter and the engine sensors measure the measurable parameters. The OL-SW-DNN, 20 which has stronger fitting capacity than other shallow network structure such as traditional NN 21 and suppose vector regression, 22 is adopted to fitting the transient and steady process of engine. The SVM can be calculated through linearizing the OL-SW-DNN model and chosen as predictive model.…”

Section: Introductionmentioning

confidence: 99%

Aero-engine direct thrust control with nonlinear model predictive control based on linearized deep neural network predictor

Proceedings of the Institution of Mechanical Engineers, Part I:

Yong

Sun

et al. 2019

Self Cite

A novel nonlinear model predictive control method for aero-engine direct thrust control is proposed to improve engine response ability and reduce computational complexity of nonlinear model predictive control. The control objective of the proposed method is the thrust directly instead of the measurable parameters. The linearized model based on online sliding window deep neural network is proposed as predictive model. The online sliding window deep neural network has strong fitting capacity for nonlinear object and adopted to fitting the transient process of engine. The back propagation is adopted to obtain linearized model of online sliding window deep neural network, which greatly reduce the calculated amount. The comparison simulations of the popular nonlinear model predictive control based on extended Kalman filter and the proposed one are carried out. The simulation results show that compared with the popular nonlinear model predictive control, the proposed nonlinear model predictive control not only has the better response ability but also has reduced computational complexity greatly, nearly reduce computation time more than 35 ms.

“…The CLM has better accuracy, but it has poor real-time performance. Therefore, many scholars proposed support vector machine (SVM) [19][20][21] and neural network (NN) 22,23 for engine modeling. The accuracy and real time of these two methods are a compromise between the PLM and the CLM.…”

Section: Introductionmentioning

confidence: 99%

A study on global optimization and deep neural network modeling method in performance-seeking control

Proceedings of the Institution of Mechanical Engineers, Part I:

Wang

et al. 2019

Self Cite

In this article, a novel performance-seeking control method based on deep neural network and interval analysis is proposed to obtain a better engine performance. A deep neural network modeling method which has stronger representation capability than conventional neural network and can deal with big training data is adopted to establish an on-board model in the subsonic and supersonic cruising envelops. Meanwhile, a global optimization algorithm interval analysis is applied here to get a better engine performance. Finally, two simulation experiments are conducted to verify the effectiveness of the proposed methods. One is the on-board model modeling which compares the deep neural network with the conventional neural network, and the other is the performance-seeking control simulations comparing interval analysis with feasible sequential quadratic programming, particle swarm optimization, and genetic algorithm, respectively. These two experiments show that the deep neural network has much higher precision than the conventional neural network and the interval analysis gets much better engine performance than feasible sequential quadratic programming, particle swarm optimization, and genetic algorithm.