A Novel Performance Adaptation and Diagnostic Method for Aero-Engines Based on the Aerothermodynamic Inverse Model

Aero-engines are faced with severe challenges of availability and reliability in the increasing operation, and traditional gas path filtering diagnostic methods have limitations restricted by various factors such as strong nonlinearity of the system and lack of critical sensor information. A method based on the aerothermodynamic inverse model (AIM) is proposed to improve the adaptation accuracy and fault diagnostic dynamic estimation response speed in this paper. Thermodynamic mechanisms are utilized to develop AIM, and scaling factors are designed to be calculated iteratively in the presence of measurement correction. In addition, the proposed method is implemented in combination with compensation of the nonlinear filter for real-time estimation of health parameters under the hypothesis of estimated dimensionality reduction. Simulations involved experimental datasets revealed that the maximum average simulated error decreased from 13.73% to 0.46% through adaptation. It was also shown that the dynamic estimated convergence time of the improved diagnostic method reached 2.183 s decrease averagely without divergence compared to the traditional diagnostic method. This paper demonstrates the proposed method has the capacity to generalize aero-engine adaptation approaches and to achieve unbiased estimation with fast convergence in performance diagnostic techniques.

An Effective Antenna Array Diagnosis Method via Multivalued Neural Network Inverse Modeling Approach

Failure of element (s) in antenna arrays impair (s) symmetry and lead to unwanted distorted radiation pattern. The replacement of defective elements in aircraft antennas is a solution to the problem, but it remains a critical problem in space stations. In this paper, an antenna array diagnosis technique based on multivalued neural network (mNN) inverse modeling is proposed. Since inverse analytical input-to-output formulation is generally a challenging and important task in solving the inverse problem of array diagnosis, ANN is a compelling alternative, because it is trainable and learns from data in inverse modelling. The mNN technique proposed is an inverse modelling technique, which accommodates measurements for output model. This network takes radiation pattern samples with faults and matches it to the corresponding position or location of the faulty elements in that antenna array. In addition, we develop a new training error function, which focuses on the matching of each training sample by a value of our proposed inverse model, while the remaining values are free, and trained to match distorted radiation patterns. Thereby, mNN learns all training data by redirecting the faulty elements patterns into various values of the inverse model. Therefore, mNN is able to perform accurate array diagnosis in an automated and simpler manner.

A grey hysteresis model of magnetorheological damper

Owing to the complex nonlinear hysteresis of magnetorheological (MR) damper, the modeling of an MR damper is an issue. This paper examines a novel MR damper hysteresis model based on the grey theory, which can fully mine the internal laws for the data with small samples and poor information. To validate the model, the experiment is conducted in the MTS platform, and then the experimental results are compiled to identify the model parameters. Considering the complexity of the grey model and its inverse model solution, the grey model is simplified in two ways based on the grey relational analysis method. Furthermore, the simplified grey model compares to other models to prove the superiority of the grey model. The analysis suggests the fitting results correspond to the measured results, and the mean relative error (MRE) of grey model is within 2.04%. After the grey model is simplified, its accuracy is slightly reduced, while its inverse model is easier to solve and makes a unique solution. Finally, compared with the polynomial and Bouc-Wen model, the novel model with fewer identification parameters has high accuracy and predictive ability. This novel model has fabulous potential in designing the control strategy of MR damper.

From internal models toward metacognitive AI

In several papers published in Biological Cybernetics in the 1980s and 1990s, Kawato and colleagues proposed computational models explaining how internal models are acquired in the cerebellum. These models were later supported by neurophysiological experiments using monkeys and neuroimaging experiments involving humans. These early studies influenced neuroscience from basic, sensory-motor control to higher cognitive functions. One of the most perplexing enigmas related to internal models is to understand the neural mechanisms that enable animals to learn large-dimensional problems with so few trials. Consciousness and metacognition—the ability to monitor one’s own thoughts, may be part of the solution to this enigma. Based on literature reviews of the past 20 years, here we propose a computational neuroscience model of metacognition. The model comprises a modular hierarchical reinforcement-learning architecture of parallel and layered, generative-inverse model pairs. In the prefrontal cortex, a distributed executive network called the “cognitive reality monitoring network” (CRMN) orchestrates conscious involvement of generative-inverse model pairs in perception and action. Based on mismatches between computations by generative and inverse models, as well as reward prediction errors, CRMN computes a “responsibility signal” that gates selection and learning of pairs in perception, action, and reinforcement learning. A high responsibility signal is given to the pairs that best capture the external world, that are competent in movements (small mismatch), and that are capable of reinforcement learning (small reward-prediction error). CRMN selects pairs with higher responsibility signals as objects of metacognition, and consciousness is determined by the entropy of responsibility signals across all pairs. This model could lead to new-generation AI, which exhibits metacognition, consciousness, dimension reduction, selection of modules and corresponding representations, and learning from small samples. It may also lead to the development of a new scientific paradigm that enables the causal study of consciousness by combining CRMN and decoded neurofeedback.

Contribution of the generalized nonunique inverses to the minimum-energy control theory: The inverse model control investigation

The innovative analytical approach to the minimum-energy design problem of the inverse model control (IMC) state-space structures is presented in this work. Following the recent papers, it should be concluded that the optimal behavior of the IMC strategy cannot longer be associated with the application of the well-known Moore–Penrose minimum-norm inverse. However, the minimum-energy IMC-oriented scheme has only be obtained through heuristic methods. Nevertheless, in the recent authors’ work, it has been proven for the first time that such an issue can be considered in an analytical manner. Yet, the obtained results have only been valid for the second-order state-space systems. Therefore, the motivation instance proposed in the manuscript, confirming the possibility of extending such paradigm to higher-order plants, will certainly contribute to the introduction of the new unified minimum-energy IMC theory canon. Since the nonunique σ and H inverses can successfully be employed in the robustification of the discussed control strategy, they can also be helpful in the case of our essential considerations. Thus, from now on the yet unexplored research area can now be investigated in the analytical manner, what has never been seen before in the modern IMC-originated control theory and practice. The predefined methodology clearly fills the gap in the analytical control design procedures and opens a new chapter in the knowledge related to the well-known and broadly accepted multivariable control canons.