RESUMEN
Traditional signal processing methods based on acceleration signals can determine whether a fault has occurred in a planetary gearbox. However, acceleration signals are severely affected by interference, causing difficulties in fault identification. This study proposes a gear fault classification method based on root strain and pseudo images. Firstly, fiber optic sensors are employed to directly acquire strain data from the ring gear root. Next, the strain signals are preprocessed using resampling and a time-domain synchronous averaging algorithm. The processed signals are encoded into two-dimensional images using Gramian Angular Fields (GAF). Then, CN-EfficientNet with contrast learning is proposed to analyze and extract deeper fault features from the image texture features. In the classification experiments for different types of faults, the accuracy reached 96.84%. The results indicate that the method can effectively accomplish the task of fault classification in planetary gearboxes. Comparative experiments with other common classification models further indicate the superior performance of the proposed learning model. Visualization based on Grad-CAM provides interpretability for the fault recognition network's results and reveals the underlying mechanism for its excellent classification performance.
RESUMEN
Graph convolutional networks (GCNs) as the emerging neural networks have shown great success in Prognostics and Health Management because they can not only extract node features but can also mine relationship between nodes in the graph data. However, the most existing GCNs-based methods are still limited by graph quality, variable working conditions, and limited data, making them difficult to obtain remarkable performance. Therefore, it is proposed in this paper a two stage importance-aware subgraph convolutional network based on multi-source sensors named I2SGCN to address the above-mentioned limitations. In the real-world scenarios, it is found that the diagnostic performance of the most existing GCNs is commonly bounded by the graph quality because it is hard to get high quality through a single sensor. Therefore, we leveraged multi-source sensors to construct graphs that contain more fault-based information of mechanical equipment. Then, we discovered that unsupervised domain adaptation (UDA) methods only use single stage to achieve cross-domain fault diagnosis and ignore more refined feature extraction, which can make the representations contained in the features inadequate. Hence, it is proposed the two-stage fault diagnosis in the whole framework to achieve UDA. In the first stage, the multiple-instance learning is adopted to obtain the importance factor of each sensor towards preliminary fault diagnosis. In the second stage, it is proposed I2SGCN to achieve refined cross-domain fault diagnosis. Moreover, we observed that deficient and limited data may cause label bias and biased training, leading to reduced generalization capacity of the proposed method. Therefore, we constructed the feature-based graph and importance-based graph to jointly mine more effective relationship and then presented a subgraph learning strategy, which not only enriches sufficient and complementary features but also regularizes the training. Comprehensive experiments conducted on four case studies demonstrate the effectiveness and superiority of the proposed method for cross-domain fault diagnosis, which outperforms the state-of-the art methods.
RESUMEN
Dental implants have seen widespread and successful use in recent years. Given their long-term application and the critical role of geometry in determining fracture and fatigue characteristics, fatigue assessments are of utmost importance for implant systems. In this study, nine dental implant system samples were subjected to testing in accordance with ISO 14801 standards. The tests included static evaluations to assess ultimate loads and fatigue tests conducted under loads of 270 N and 230 N at a frequency of 15 Hz, aimed at identifying fatigue failure locations and fatigue life. Fatigue life predictions and related calculations were carried out using Fe-safe software. The initial model featured a 22° angle for both the fixture and abutment. Subsequently, variations in abutment angles at 21° and 23° were considered while keeping the fixture angle at 22°. In the next phase, the fixture and abutment angles were set as identical, at 21° and 23°. The results unveiled that when the angles of the abutment and fixture matched, stress values decreased, and fatigue life increased. Conversely, models featuring abutment angles of 21° and 23°, with a 22° angle for the fixture, led to a 49.1 % increase in stress and a 36.9 % decrease in fatigue life compared to the primary model. Notably, in the case of the implant with a 23° angle for both abutment and fixture, the fatigue life reached its highest value at 10 million cycles. Conversely, the worst-case scenario was observed in the implant with a 21° abutment angle and a 23° fixture angle, with a fatigue life of 5.49 million cycles.
Asunto(s)
Implantes Dentales , Análisis del Estrés Dental , Estrés MecánicoRESUMEN
Recently, due to the difficulty of collecting condition data covering all mechanical fault types in industrial scenarios, the fault diagnosis problem under incomplete data is receiving increasing attention where no target prior information can be available. The existing open-set or universal domain adaptation (DA) diagnosis methods typically treat private fault samples in the target as a generalized "unknown" fault class, neglecting their inherent structure. This oversight can lead to confusion in latent feature space representations and difficulties in separating unknown samples. Therefore, a universal DA method with unsupervised clustering is developed to explore the intrinsic structure of the target samples for mechanical fault diagnosis, where multi-source information on different working conditions is considered to transfer complementary knowledge. First, a composite clustering metric combining single-domain and cross-domain evaluation is constructed to recognize shared and unknown health classes on source-target domains. Second, to alleviate the intra-class shift while enlarging the inter-class gap, a class-wise DA algorithm is suggested which operates on the basis of maximum mean discrepancy. Finally, an entropy regularization criterion is utilized to facilitate clustering of different health classes. The efficacy of the presented approach in the fault diagnosis issues when monitoring data is inadequate has been verified through extensive experiments on three rotating machinery datasets.
Asunto(s)
Algoritmos , Conocimiento , Análisis por Conglomerados , EntropíaRESUMEN
This paper aims to study the fixed-time stabilization of a class of delayed discontinuous reaction-diffusion Cohen-Grossberg neural networks. Firstly, by providing some relaxed conditions containing indefinite functions and based on inequality techniques, a new fixed-time stability lemma is given, which can improve the traditional ones. Secondly, based on state-dependent switching laws, the periodic wave solution of the formulated networks is transformed into the periodic solution of ordinary differential system. By utilizing differential inclusions theory and coincidence theorem, the existence of periodic solutions is obtained. Thirdly, based on the new fixed-time stability lemma, the periodic solutions are stabilized at zero in a fixed-time, which is a new topic on reaction-diffusion networks. Moreover, the established criteria are all delay-dependent, which are less conservative than the previous delay-independent ones for ensuring the stabilization of delayed reaction-diffusion networks. Finally, two examples give numerical explanations of the proposed results and highlight the influence of delays.
Asunto(s)
Algoritmos , Redes Neurales de la Computación , Factores de TiempoRESUMEN
This paper is devoted to the issue of observer-based adaptive sliding mode control of distributed delay systems with deterministic switching rules and stochastic jumping process, simultaneously, through a neural network approach. Firstly, relying on the designed Lebesgue observer, a sliding mode hyperplane in the integral form is put forward, on which a desired sliding mode dynamic system is derived. Secondly, in consideration of complexity of real transition rates information, a novel adaptive dynamic controller that fits to universal mode information is designed to ensure the existence of sliding motion in finite-time, especially for the case that the mode information is totally unknown. In addition, an observer-based neural compensator is developed to attenuate the effectiveness of unknown system nonlinearity. Thirdly, an average dwell-time approach is utilized to check the mean-square exponential stability of the obtained sliding mode dynamics, particularly, the proposed criteria conditions are successfully unified with the designed controller in the type of mode information. Finally, a practical example is provided to verify the validity of the proposed method.
Asunto(s)
Redes de Comunicación de Computadores , Redes Neurales de la Computación , Movimiento (Física)RESUMEN
Playing games between humans and robots have become a widespread human-robot confrontation (HRC) application. Although many approaches were proposed to enhance the tracking accuracy by combining different information, the problems of the intelligence degree of the robot and the anti-interference ability of the motion capture system still need to be solved. In this paper, we present an adaptive reinforcement learning (RL) based multimodal data fusion (AdaRL-MDF) framework teaching the robot hand to play Rock-Paper-Scissors (RPS) game with humans. It includes an adaptive learning mechanism to update the ensemble classifier, an RL model providing intellectual wisdom to the robot, and a multimodal data fusion structure offering resistance to interference. The corresponding experiments prove the mentioned functions of the AdaRL-MDF model. The comparison accuracy and computational time show the high performance of the ensemble model by combining k-nearest neighbor (k-NN) and deep convolutional neural network (DCNN). In addition, the depth vision-based k-NN classifier obtains a 100% identification accuracy so that the predicted gestures can be regarded as the real value. The demonstration illustrates the real possibility of HRC application. The theory involved in this model provides the possibility of developing HRC intelligence.
Asunto(s)
Robótica , Juegos de Video , Humanos , Refuerzo en Psicología , Redes Neurales de la Computación , AprendizajeRESUMEN
Intelligent fault diagnosis aims to build robust mechanical condition recognition models with limited dataset. At this stage, fault diagnosis faces two practical challenges: (1) the variability of mechanical working conditions makes the collected data distribution inconsistent, which brings about the domain shift; (2) some unpredictable unknown fault modes that do not observe in the training dataset may occur in the testing scenario, leading to a category gap. In order to cope with these two entangled challenges, an open set multi-source domain adaptation approach is developed in this study. Specifically, a complementary transferability metric defined on multiple classifiers is introduced to quantify the similarity of each target sample to known classes to weight the adversarial mechanism. By applying an unknown mode detector, unknown faults can be automatically identified. Moreover, a multi-source mutual-supervised strategy is further adopted to mine relevant information between different sources to enhance the model performance. Extensive experiments are conducted on three rotating machinery datasets, and the results show that the proposed method is superior to traditional domain adaptation approaches in the mechanical diagnosis issues that new fault modes occur.
Asunto(s)
Aprendizaje Profundo , Recolección de Datos , Inteligencia , Reconocimiento en Psicología , Condiciones de TrabajoRESUMEN
In the context of motion planning in robotics, the problem of path planning based on artificial potential fields has been examined using different algorithms to avoid trapping in local minima. With this objective, this paper proposes a novel method based on a deterministic annealing strategy to improve the potential field function by introducing a temperature parameter to increase the robot's obstacle avoidance efficiency. The annealing and tempering strategies prevent the robot from being trapped at the local minima and allow it to continue towards its destination. The initial path is optimised using an annealing algorithm to enhance the overall performance. The time, length and success rate of the planned path measures the quality of the solution. Simulation results and comparative experiments demonstrate that the proposed algorithm can solve path planning in different environments. The proposed algorithm is suitable for complex environments with convex or non-convex polygon obstacles.
RESUMEN
This paper is concerned with the measurement outlier-resistant mobile robot localization problem by using multiple Doppler-azimuth radars under round-robin protocol (R-RP). In the considered robot localization system, multiple Doppler-azimuth radars are equipped on the robot platform to produce the measurement including the Doppler frequency shift and the azimuth. In order to assuage communication link congestion, the R-RP is used. For mitigating the influence of outliers, a time-varying state estimator is constructed which contains a saturation function with variable saturation levels. This paper aims at seeking out a practicable yet effective solution to the addressed robot localization problem by devising the constructed estimator which can assure that, over a finite horizon, the localization error satisfies the given H∞ performance index. By constructing an appropriate Lyapunov function, the sufficient condition, which can guarantee the localization error to fulfill the given H∞ performance, is established. Then, by resorting to the solution to a set of linear matrix inequalities, the constructed estimator can be devised. In the light of the estimator design strategy proposed in this paper, the corresponding robot localization algorithm is developed. At last, some simulations are conducted to testify the usefulness of the developed robot localization algorithm.
RESUMEN
This paper is dealing with the problem of observer-based event-triggered sliding mode control for fractional-order uncertain switched systems with a positive order less than one. Firstly, a fractional-order state observer is designed, based on which a fractional-order integral sliding surface function is proposed. Then, utilizing the estimated observer error and sliding mode error vectors, an event-triggered condition is constructed to decide whether the current control signal should be updated or not. Besides, sufficient conditions are derived in the forms of linear matrix inequalities (LMIs) to ensure finite-time stability of the augmented closed-loop system by adopting an average dwell time approach. Thereafter, to avoid the occurrence of infinite triggers within finite time, this paper also discusses the Zeno behavior and refines the results in the previous literature. Finally, to illustrate the effectiveness and superiority of the proposed method, three numerical simulations are provided.
RESUMEN
The state-of-energy (SOE) and state-of-health (SOH) are two crucial quotas in the battery management systems, whose accurate estimation is facing challenges by electric vehicles' (EVs) complexity and changeable external environment. Although the machine learning algorithm can significantly improve the accuracy of battery estimation, it cannot be performed on the vehicle control unit as it requires a large amount of data and computing power. This paper proposes a joint SOE and SOH prediction algorithm, which combines long short-term memory (LSTM), Bi-directional LSTM (Bi-LSTM), and convolutional neural networks (CNNs) for EVs based on vehicle-cloud collaboration. Firstly, the indicator of battery performance degradation is extracted for SOH prediction according to the historical data; the Bayesian optimization approach is applied to the SOH prediction combined with Bi-LSTM. Then, the CNN-LSTM is implemented to provide direct and nonlinear mapping models for SOE. These direct mapping models avoid parameter identification and updating, which are applicable in cases with complex operating conditions. Finally, the SOH correction in SOE estimation achieves the joint estimation with different time scales. With the validation of the National Aeronautics and Space Administration battery data set, as well as the established battery platform, the error of the proposed method is kept within 3%. The proposed vehicle-cloud approach performs high-precision joint estimation of battery SOE and SOH. It can not only use the battery historical data of the cloud platform to predict the SOH but also correct the SOE according to the predicted value of the SOH. The feasibility of vehicle-cloud collaboration is promising in future battery management systems.
Asunto(s)
Suministros de Energía Eléctrica , Electricidad , Estados Unidos , Teorema de Bayes , Fenómenos Físicos , Redes Neurales de la ComputaciónRESUMEN
This paper is devoted to design an event-triggered data-driven control for a class of disturbed nonlinear systems with quantized input. A uniform quantizer reconstructed with decreasing quantization intervals is employed to reduce the quantization error. A neural network-based estimation strategy is proposed to estimate both the pseudo partial derivative and disturbances. Consequently, an input triggering rule for single-input single-output systems is provided by incorporating the estimated disturbances, the quantization error bound and tracking errors. Resorting to the Lyapunov method, sufficient conditions for synthesized error systems to be uniformly ultimately bounded are presented. The validity of the proposed scheme is demonstrated via a simulation example.
Asunto(s)
Redes Neurales de la Computación , Dinámicas no Lineales , Simulación por ComputadorRESUMEN
This paper is devoted to studying the passivity-based sliding mode control for nonlinear systems and its application to dock cranes through an adaptive neural network approach, where the system suffers from time-varying delay, external disturbance and unknown nonlinearity. First, relying on the generalized Lagrange formula, the mathematical model for the crane system is established. Second, by virtue of an integral-type sliding surface function and the equivalent control theory, a sliding mode dynamic system can be obtained with a satisfactory dynamic property. Third, based on the RBF neural network approach, an adaptive control law is designed to ensure the finite-time existence of sliding motion in the face of unknown nonlinearity. Fourth, feasible easy-checking linear matrix inequality conditions are developed to analyze passification performance of the resulting sliding motion. Finally, a simulation study is provided to confirm the validity of the proposed method.
Asunto(s)
Algoritmos , Dinámicas no Lineales , Simulación por Computador , Modelos Teóricos , Redes Neurales de la ComputaciónRESUMEN
This paper uses numerical approach to give insight into the structural-acoustic response of a lightweight square aluminium panel. It takes into consideration different locations of a primary sound source in an acoustic medium and how these locations influence the response of the structural panel. Finite element method as well as the first-order deformation theory are employed for constructing the numerical model. Experimental measurements of the mode shapes and velocity frequency response of the vibrating panel are used to validate the results of the finite element model. Furthermore, vibro-acoustic emission indexes such as sound transmission loss, sound pressure level and far-field directivity of sound pressure are obtained numerically. The results show that different locations of the primary sound source significantly influence the response of the structural panel to reduce noise. Sound source typically positioned close to the structural panel lowers the efficiency of the vibrating panel to reduce noise. Moreover, the sound distribution profiles at the radiated end of the vibrating panel for the different locations of the sound source are investigated. The study shows that the variation of the zones of quiet, vibro-acoustic emission parameters and sound distribution profiles obtained can provide vital information about the best positioning of structural source for both active vibration and noise control.
RESUMEN
Blood pressure (BP) is known as an indicator of human health status, and regular measurement is helpful for early detection of cardiovascular diseases. Traditional techniques for measuring BP are either invasive or cuff-based and thus are not suitable for continuous measurement. Aiming at the deficiencies in existing studies, a novel cuffless BP estimation framework of Receptive Field Parallel Attention Shrinkage Network (RFPASN) and BP range constraint is proposed. Firstly, RFPASN uses the multi-scale large receptive field convolution module to capture the long-term dynamics in the photoplethysmography (PPG) signal without using long short-term memory (LSTM). On this basis, the features acquired by the parallel mixed domain attention module are used as thresholds, and the soft threshold function is used to screen the input features to enhance the discriminability and robustness of features, which can significantly improve the prediction accuracy of diastolic blood pressure (DBP) and systolic blood pressure (SBP). Finally, in order to prevent large fluctuations in the prediction results of RFPASN, RFPASN based on BP range constraint is proposed to make the prediction results of RFPASN more accurate and reasonable. The performance of the proposed method is demonstrated on a publically available MIMIC-II database. The database contains normal, hypertensive and hypotensive people. We have achieved MAE of 1.63/1.59 (DBP) and 2.26/2.15 (SBP) mmHg for BP on total population of 1562 subjects. A comparative study shows that the proposed algorithm is more promising than the state-of-the-art.
Asunto(s)
Determinación de la Presión Sanguínea , Aprendizaje Profundo , Algoritmos , Presión Sanguínea/fisiología , Determinación de la Presión Sanguínea/métodos , Humanos , Fotopletismografía/métodosRESUMEN
Some artificial intelligence algorithms have gained much attention in the rotating machinery fault diagnosis due to their robust nonlinear regression properties. In addition, existing deep learning algorithms are usually dependent on single signal features, which would lead to the loss of some information or incomplete use of the information in the signal. To address this problem, three kinds of popular signal processing methods, including Fast Fourier Transform (FFT), Short-Time Fourier Transform (STFT) and directly slicing one-dimensional data into the two-dimensional matrix, are used to create four different datasets from raw vibration signal as the input data of four enhancement Convolutional Neural Networks (CNN) models. Then, a fuzzy fusion strategy is used to fuse the output of four CNN models that could analyze the importance of each classifier and explore the interaction index between each classifier, which is different from conventional fusion strategies. To show the performance of the proposed model, an artificial fault bearing dataset and a real-world bearing dataset are used to test the feature extraction capability of the model. The good anti-noise and interpretation characteristics of the proposed method are demonstrated as well.
Asunto(s)
Inteligencia Artificial , Redes Neurales de la Computación , Algoritmos , Procesamiento de Señales Asistido por Computador , VibraciónRESUMEN
This paper investigates the problem of output feedback neural network (NN) learning tracking control for nonlinear strict feedback systems subject to prescribed performance and input dead-zone constraints. First, an NN is utilized to approximate the unknown nonlinear functions, then a state observer is developed to estimate the unmeasurable states. Second, based on the command filter method, an output feedback NN learning backstepping control algorithm is established. Third, a prescribed performance function is employed to ensure the transient performance of the closed-loop systems and forces the tracking error to fall within the prescribed performance boundary. It is rigorously proved mathematically that all the signals in the closed-loop systems are semi-globally uniformly ultimately bounded and the tracking error can converge to an arbitrarily small neighborhood of the origin. Finally, a numerical example and an application example of the electromechanical system are given to show effectiveness of the acquired control algorithm.
Asunto(s)
Redes Neurales de la Computación , Dinámicas no Lineales , Algoritmos , Simulación por Computador , RetroalimentaciónRESUMEN
This article studies the lag-bipartite formation tracking (LBFT) problem of the networked robotic systems (NRSs) with directed matrix-weighted signed graphs. Unlike the traditional formation tracking problems with only cooperative interactions, solving the LBFT problem implies that: 1) the robots of the NRS are divided into two complementary subgroups according to the signed graph, describing the coexistence of cooperative and antagonistic interactions; 2) the states of each subgroup form a desired geometric pattern asymptotically in the local coordinate; and 3) the geometric center of each subgroup is forced to track the same leader trajectory with different plus-minus signs and a time lag. A new hierarchical control algorithm is designed to address this challenging problem. Based on the Lyapunov stability argument and the property of the matrix-weighted Laplacian, some sufficient criteria are derived for solving the LBFT problem. Finally, simulation examples are proposed to validate the effectiveness of the main results.