A New High-Precision General Inverse Model Based on Deep Learning for Magnetorheological Damper

Magnetorheological (MR) damper is a prominent semi-active control device for earthquake response mitigation of structures. The most important topic for the intelligent MR structures is choosing the control current of MR damper quickly and accurately. The typical control strategy is on-off control strategy, while inherent time-delay and coarse control precision lie in this strategy. This article proposes a semi-active control strategy for restricting the MR structures efficiently, in which the genetic algorithm (GA) is adopted to search for the optimal feedback control gain in time-delayed dynamic equations forming a GA controller, and next optimize the control current of the MR damper accurately by a series of GA procedure. The GA strategy has the capability to effectively control the MR structure involving time delay and adjusting for controlled current of the MR damper timely and accurately. Accordingly, numerical examples are used to testify that the feedback gain optimized by GA has the capability to reduce a given objective response dramatically. In addition, a comparison between GA controller and LQR controller indicates that the proposed approach performs much better than LQR method, especially in time-delay systems. Finally, simulation results for a seismically excited structure subjected by the MR damper are presented to verify that the semi-active strategy using GA is predominant in mitigating the dynamic structure efficiently. The strategy and the results presented will also throw light on the development and control of other novel structures.

This paper presents an approach for learning the Takagi-Sugeno (T-S) fuzzy model by Genetic Algorithm (GA). In this approach, the fuzzy rule structure is encoded by binary code in the chromosome in which the position of 1 indicates the selected rules and the sum of 1 indicates the number of rules. The membership function (MF) parameters (centres and bases) are evolved by GA in combining with the pseudo-inversion algorithm for obtaining the consequent parameters. The sum of squared error (SSE) between the true output and the T-S model prediction is used as objective function. Then, this approach is applied to the modelling of dynamic behaviour of a magneto-rheological (MR) damper which shows highly nonlinear characteristics due to hysteretic phenomenon. It is shown by the validation test that the developed T-S fuzzy model can represent the dynamic behaviour of the MR damper satisfactorily.

It is considerable to solve practical fault diagnosis task of gearbox under variable working conditions by introducing sufficient auxiliary data. For this purpose, a new approach called improved deep transfer auto-encoder is proposed for intelligent diagnosis of gearbox faults under variable working conditions with small training samples. First, multi-wavelet is employed as activation function for effectively learning useful features hidden in the non-stationary vibration data. Second, correntropy is used to modify the cost function to enhance the reconstruction quality. Third, pre-train an improved deep auto-encoder using sufficient auxiliary data in the source domain, and transfer its parameters to the target model. Finally, the improved deep transfer should be fine-tuned by small training samples in the target domain to adapt to the characteristics of the rest testing data. The proposed approach is used to analyze two sets of experimental vibration data collected from gearbox under variable working conditions. The results show that the proposed approach can accurately diagnose different faults of gearbox even the working conditions have significant changes, which is superior to the existing methods.

Optimum semi-active control with a limited number of magneto-rheological (MR) dampers and measurement sensors has certain requirements. Most important of them is the accurate estimation of control forces developed in the MR dampers from the observations made in the structure. Therefore, the observation strategy should form an integral part of the optimization problem. The existing literature on the subject does not address this issue properly. The paper presents a computationally efficient optimization scheme for semi-active control of partially observed building frames using a limited number of MR dampers and sensors for earthquakes. The control scheme duly incorporates the locations of measurement sensors as variables into the genetic algorithm (GA) based optimization problem. A ten-storied building frame is taken as an illustrative example. The optimum control strategy utilizes two well-known control laws, namely, the linear quadratic Gaussian (LQG) with clipped optimal control and the bang-bang control to find the time histories of voltage to be applied to the MR dampers. The results of the numerical study show that the proposed scheme of sensor placement provides the optimum reduction of response with more computational efficiency. Second, optimal locations of sensors vary with the response quantities to be controlled, the nature of earthquake, and the control algorithm. Third, optimal locations of MR dampers are invariant of the response quantities to be controlled and the nature of earthquake.

In order to address the issue of reduced damping force dynamic range in magnetorheological (MR) damper caused by the high zero-field viscosity of MR grease, known for its sedimentation stability, this paper introduces a novel dual-channel independent-coil MR damper (DCICMRD). Firstly, the dual-channel configuration and the magnetic circuit structure of independent coils were meticulously designed, and a genetic algorithm was employed to conduct multi-objective optimization of the structural parameters for DCICMRD. The optimization results indicate that all performance metrics of the damper post-optimization exhibit improvements exceeding 15%. Then, the theocratical model of the damping force for DCICMRD under three operational modes are established, and the output damping force of various input currents for different operating mode is obtained. Finally, the DCICMRD was manufactured and subjected to dynamic performance testing. The results revealed that the damping force dynamic range in Mode III, i.e. dual-channel structure, can achieve approximately 15 times, whereas Mode I, i.e. traditional single-channel structure, only attains approximately 9 times. The aforementioned research holds significant implications for expanding the further engineering applications of MR dampers.

Fault diagnosis plays a very important role in ensuring the safe and reliable operations of machines. Currently, the deep learning-based fault diagnosis is attracting increasing attention. However, fault diagnosis under variable working conditions has been a significant challenge due to the domain discrepancy problem. This problem is also unavoidable in deep learning-based fault diagnosis methods. This paper contributes to the ongoing investigation by proposing a new approach for the fault diagnosis under variable working conditions based on STFT and transfer deep residual network (TDRN). The STFT was employed to convert vibration signal to time-frequency image as the input of the TDRN. To address the domain discrepancy problem, the TDRN was developed in this paper. Unlike traditional deep convolutional neural network (DCNN) methods, by combining with transfer learning, the TDRN can make a bridge between two different working conditions, thereby using the knowledge learned from a working condition to achieve a high classification accuracy in another working condition. Moreover, since the residual learning is introducing, the TDRN can overcome the problems of training difficulty and performance degradation existing in traditional DCNN methods, thus further improving the classification accuracy. Experiments are conducted on the popular CWRU bearing dataset to validate the effectiveness and superiority of the proposed approach. The results show that the developed TDRN outperforms those methods without transfer learning and/or residual learning in terms of the accuracy and feature learning ability for the fault diagnosis under variable working conditions.

Driver tiredness during long term travel might result in many unwilling accidents. Hence, isolating the vibration transmitted to the driver plays a major role in keeping the driver restful and consequently having a safe trip. In this study, the transmitted vibration of a wagon suspension to a passenger model is isolated by a semi active control strategy using a magneto rheological (MR) damper. First, a revised linear lumped parameter biomechanical model of seated human body exposed to vertical vibrations is optimized. To this end, the parameters of the human body model, including mass, damper and stiffness coefficients are simultaneously optimized using genetic algorithm by employing a multi-objective function. Next, the optimized human body model is integrated on a nonlinear seat suspension model. This integrated model is excited by a sinusoidal frequency range as applicable to railway vehicles. The responses in time and frequency domains are derived. Finally, by using a MR damper instead of conventional dampers in the seat suspension, it is shown that the transmitted acceleration from the rail to the driver is reduced. Finally, a comparison of the performances of passive conventional dampers, semi-active MR dampers and active actuator assisted dampers is presented in detail.

Integrated fuzzy logic and genetic algorithms for multi-objective control of structures using MR dampers

Machine learning methods have made great development in data-driven fault diagnosis of rolling bearings. But the intelligent fault diagnosis of intershaft bearing faces the following two dilemmas: 1) the fault vibration is extremely weak, and it is difficult to extract features that can distinguish different classes; 2) due to the complex and variable working condition, the intershaft bearing does not always fail in same working conditions. That is, the labeled training data is not sufficient for every source domain. These challenges lead to the failure of traditional machine learning based fault diagnosis for intershaft bearings. Therefore, a novel intelligent fault diagnosis scheme is investigated for intershaft bearings of dual-rotor equipment under variable working conditions. The paper focuses on two key issues: 1) developing a feature extraction approach with which the fault features with excellent clustering and separation are extracted from vibration signals. This approach addresses the dilemma of weak fault feature extraction of intershaft bearing and creates a feasible precondition for subsequent feature transfer; 2) proposing a feature transfer method transforming the labeled sample features in multiple source domains into the trainable sample features in the target domain. This new transfer achieves the sharing of labeled training samples under working conditions and enriches the trainable samples in target domain. Ultimately, the faults of intershaft bearings can be diagnosed with the help of the neural network classifier trained by the transferred samples with labels. Experimental results verify that this established scheme is effective and superior to other comparable methods for the transfer diagnosis task from multiple source domains to target domain.

Ensuring the smooth operation of rolling bearings requires a precise fault diagnosis. Particularly, identifying fault types under varying working conditions holds significant importance in practical engineering. Thus, we propose a reinforcement ensemble method for diagnosing rolling bearing faults under varying working conditions. Firstly, a reinforcement model was designed to select the optimal base learner. Stratified random sampling was used to extract four datasets from raw training data. The reinforcement model was trained by these four datasets, respectively, and we obtained four optimal base learners. Then, a sparse ANN was designed as the ensemble model and the reinforcement learning model that can successfully identify the fault type under variable work conditions was constructed. Extensive experiments were conducted, and the results demonstrate the superiority of the proposed method over other intelligent approaches, with significant practical engineering benefits.

High-speed train brake pads state recognition faces the problems of single data source feature characterization limitation and significant domain shifts under variable working conditions. Considering that multi-source heterogeneous data can characterize the brake pad state from different physical dimensions, this paper proposes a multi-source deep adversarial decoupled autoencoder network (MS-DADA) for online identification of brake pad state of high-speed trains under variable working conditions. First, a signal characterization system covering the multidimensional state characteristics of the friction interface is constructed by fusing three kinds of multi-source heterogeneous data, including friction coefficient, tangential acceleration and noise. Second, a deep adversarial decoupled autoencoder is designed to realize the explicit decoupling of domain-invariant and domain-specific features by utilizing the synergistic mechanism of mutual information minimization constraint and domain adversarial. Finally, with the validation set accuracy as the optimization objective, a genetic algorithm is introduced to dynamically allocate multi-source weights. This adaptive weighted fusion strategy significantly enhances the model's generalization capability for unknown rotational speed conditions. The experimental results of 10 cross-speed tasks show that the proposed model achieves an average accuracy of 99.12%. It is 7.1%, 9.36% and 26.5% higher than the single-source model, and 3.58% to 6.36% better than the current leading domain generalization methods.

Characterization and quarter car analysis with magnetorheological fluid damper using modified algebraic model (mAlg)

Bearing is a key component of rotating machinery, thus the accurate and efficient fault diagnosis of bearing is critical to the safety and reliability of rotating machinery systems. In recent years, the intelligent fault diagnosis technology of bearing represented by deep learning has made rapid development. However, the stability and generalization performance of the fault diagnosis model of bearing under complex environmental noise and variable working conditions is not good. Fortunately, self-supervised learning represented by contrastive learning can provide a good solution. The current research on contrastive learning mainly focuses on self-supervised contrastive learning, but the lack of supervised learning of fault label information leads to the reduction of diagnostic accuracy and robustness under variable working conditions. To solve this problem, this paper proposes a supervised contrastive learning combined with multi-scale attention mechanism for fault diagnosis of bearing under variable operating conditions. Firstly, the preprocessing of bearing time-domain vibration data is completed, including data set division, data augmentation, and construction of “positive” or “negative” pairs for comparative learning. Then, a supervised learning network integrating multi-scale attention mechanism is constructed to complete bearing fault feature extraction. Finally, the model transfer fault diagnosis is completed under complex noise and variable working conditions. The performance of the proposed method is evaluated through a large number of noise and variable-condition experiments. The results show that the proposed method has high fault classification accuracy and robustness under the influence of complex noise and variable working conditions.

In recent years, the intelligent diagnosis of bearing fault under stable condition based on machine learning and vibration signal has made a lot of achievements. However, the performance degradation of these methods is obvious under variable working conditions, which makes the intelligent diagnosis of bearing fault still a challenge. In order to solve this problem, a rolling bearing fault transfer diagnosis method based on acoustic emission signal is proposed in this paper. The AE data under different working conditions are collected and converted to signal map firstly, then the sensitive features are extracted and the labeled knowledge in source domain is transferred to the without labeled target domain by using convolutional neural network and domain adaptation. Finally, the feasibility and superiority of the proposed method was verified by lots of bearing experiments, and the classification accuracy reached 88.40%, which is much higher than the traditional methods.

“The purpose of this study was to compare the effect of work motivation and working conditions on the performance of MSME employees before and during the Covid-19 pandemic. The researcher uses a comparative research type with a quantitative approach. Techniques Data collection is done using a questionnaire with the linkert method. And the sampling technique used is to use a saturated sample (total sampling). The total sample obtained is 37 employees from several MSMEs. In the test of the effect of work motivation, work motivation has no significant effect on employee performance. Variable working conditions have a significant effect on employee performance. In the F test there is an influence between work motivation and working conditions on employee performance. And in the comparison test, there is a significant difference between work motivation and working conditions before and during covid-19. While in the employee performance variable there is no significant difference between employee performance before and during Covid-19”.