System Identification and Parameter Self-Tuning Controller on Deep-Sea Mining Vehicle

System identification is a quintessential measure for real-time analysis on kinematic characteristics for deep-sea mining vehicle, and thus to enhance the control performance and testing efficiency. In this study, the system identification algorithm, recursive least square method with instrumental variables(IV-RLS), is tailored to model ‘Pioneer I’, a deep-sea mining vehicle which recently completed a 1305-meter-deep sea trial in the Xisha area of the South China Sea in August, 2021. The algorithm operates on the sensor data collected from the trial to obtain the vehicle’s kinematic model and accordingly design the parameter self-tuning controller. The performances demonstrate the accuracy of the model, and prove its generalization capability. With this model, the optimal controller has been designed, the control parameters have been self-tuned, and the response time and robustness of the system have been optimized,which validates the high efficiency on digital modelling for precision control of deep-sea mining vehicles.

Nonlinear Dynamic System Identification of ARX Model for Speech Signal Identification

System Identification becomes very crucial in the field of nonlinear and dynamic systems or practical systems.As most practical systems don’t have prior information about the system behaviour thus,mathematical modelling is required.The authors have proposed a stacked Bidirectional Long-Short Term Memory(Bi-LSTM)model to handle the problem of nonlinear dynamic system identification in this paper.The proposed model has the ability of faster learning and accurate modelling as it can be trained in both forward and backward directions.The main advantage of Bi-LSTM over other algorithms is that it processes inputs in two ways:one from the past to the future,and the other from the future to the past.In this proposed model a backward-running Long-Short Term Memory(LSTM)can store information from the future along with application of two hidden states together allows for storing information from the past and future at any moment in time.The proposed model is tested with a recorded speech signal to prove its superiority with the performance being evaluated through Mean Square Error(MSE)and Root Means Square Error(RMSE).The RMSE and MSE performances obtained by the proposed model are found to be 0.0218 and 0.0162 respectively for 500 Epochs.The comparison of results and further analysis illustrates that the proposed model achieves better performance over other models and can obtain higher prediction accuracy along with faster convergence speed.

IMPROVEMENT ACCURACY OF SYSTEM IDENTIFICATION BASED ON MORLET WAVELET

A time frequency de-noising method is presented in the frequency response function （FRF） preprocessing based on the continuous wavelet transform. Morlet wavelet is employed to construct a filter bank to reduce the noise. The filter bank is a finite impulse response （FIR） linear phase filter thus maintaining phase consistency. A modified Morlet base function is proposed to meet the time frequency resolution by using transient excitation. Numerical simulation is conducted using a Group for Aeronautical Research and Technology in Europe （GARTEUR） aircraft model excited by the transient input. The white noise is added to the simulated data. Results show that the accuracy of the system identification is improved. The estimated error of the mode damping is decreased by 30% compared with that obtained from the noise-corrupted signal.

System Identification Modeling of Ship Manoeuvring Motion in 4 Degrees of Freedom Based on Support Vector Machines

Based on support vector machines, three modeling methods, i.e., white-box modeling, grey-box modeling and black-box modeling of ship manoeuvring motion in 4 degrees of freedom are investigated. With the whole-ship mathematical model for ship manoeuvring motion, in which the hydrodynamic coefficients are obtained from roll planar motion mechanism test, some zigzag tests and turning circle manoeuvres are simulated. In the white-box modeling and grey-box modeling, the training data taken every 5 s from the simulated 20°/20° zigzag test are used, while in the black-box modeling, the training data taken every 5 s from the simulated 15°/15°, 20°/20° zigzag tests and 15°, 25° turning manoeuvres are used; and the trained support vector machines are used to predict the whole 20°/20° zigzag test. Comparisons between the simulated and predicted 20°/20° zigzag tests show good predictive ability of the proposed methods. Besides, all mathematical models obtained by the proposed modeling methods are used to predict the 10°/10° zigzag test and 35° turning circle manoeuvre, and the predicted results are compared with those of simulation tests to demonstrate the good generalization performance of the mathematical models. Finally, the proposed modeling methods are analyzed and compared with each other in aspects of application conditions, prediction accuracy and computation speed. The appropriate modeling method can be chosen according to the intended use of the mathematical models and the available data needed for system identification.

Study of Stochastic Subspace System Identification Method

Stochastic Subspace Identification (SSI) is a novel time domain identification method, which directly uses operational response data to identify the system model by linear algebraic manipulations such as QR factorization and Singular Value Decomposition (SVD). This paper deals with SSI and its applications for structural modal identification. The NASA mini mast model is used for simulations to illustrate how to select input parameters, and to demonstrate identification precision. A real building structure, the Heritage Court Tower (HCT) in Canada is analyzed. From the simulation and test researches, the conclusions can be made to instruct how to identify structural modal parameters using SSI method.

Hydraulic turbine system identification and predictive control based on GASA-BPNN

Based on the characteristics of nonlinearity,multi-case,and multi-disturbance,it is difficult to establish an accurate parameter mod-el on the hydraulic turbine system which is limited by the degree of fitting between parametric model and actual model,and the design of con-trol algorithm has a certain degree of limitation.Aiming at the modeling and control problems of hydraulic turbine system,this paper proposes hydraulic turbine system identification and predictive control based on genetic algorithm-simulate anneal and back propagation neural network(GASA-BPNN),and the output value predicted by GASA-BPNN model is fed back to the nonlinear optimizer to output the control quantity.The results show that the output speed of the traditional control system increases greatly and the speed of regulation is slow,while the speed of GASA-BPNN predictive control system increases little and the regulation speed is obviously faster than that of the traditional control system.Compared with the output response of the traditional control of the hydraulic turbine governing system,the neural network predictive control-ler used in this paper has better effect and stronger robustness,solves the problem of poor generalization ability and identification accuracy of the turbine system under variable conditions,and achieves better control effect.

Experimental Study and System Identification of Hydrodynamic Force Acting on Heave Damping Plate

Although Morison equation is often applied for simulating hydrodynamic force of marine structure, it may give poor results when non-linear behavior is severe or random wave is encountered. This leads to some modifications of Morison equation or other methods for predicting hydrodynamic force. One of them is the system identification technique. In this paper, NARMAX model theory is firstly used to identify the hydrodynamic system of heave damping plates, which are commonly installed on spar platform. Both linear and non-linear models are obtained. The comparisons between the predieted results and measured data indicate that NARMAX model can predict hydrodynamic force of a heave damping plate very well. The measured data for identification originate from forced oscillation tests, which are random records with given spectrum. The forced oscillation forms in experiment also contain simple harmonic, multi-frequency ones.

System identification based on NARMAX model using Hopfield networks

An approach is proposed to avoid model structure determination in system identification using NARMAX （nonlinear autoregressive moving average with exogenous inputs） model. Identification procedure is formulated as an optimization procedure of a apecial class of Hopfield network in the proposed approach. The particular structure of these Hopfield networks can avoid the local optimum problem. Training of these Hopfield network achieves model structure determination and parameter estimation. Convergence of Hopfield networks guarantees that a NARMAX model of random initial state will approach a valid identification model with accurate state parameters. Results of two simulation examples illustrate that this approach is efficient and simple.

An Improved SPSA Algorithm for System Identification Using Fuzzy Rules for Training Neural Networks

Simultaneous perturbation stochastic approximation （SPSA） belongs to the class of gradient-free optimization methods that extract gradient information from successive objective function evaluation. This paper describes an improved SPSA algorithm, which entails fuzzy adaptive gain sequences, gradient smoothing, and a step rejection procedure to enhance convergence and stability. The proposed fuzzy adaptive simultaneous perturbation approximation （FASPA） algorithm is particularly well suited to problems involving a large number of parameters such as those encountered in nonlinear system identification using neural networks （NNs）. Accordingly, a multilayer perceptron （MLP） network with popular training algorithms was used to predicate the system response. We found that an MLP trained by FASPSA had the desired accuracy that was comparable to results obtained by traditional system identification algorithms. Simulation results for typical nonlinear systems demonstrate that the proposed NN architecture trained with FASPSA yields improved system identification as measured by reduced time of convergence and a smaller identification error.

Nonlinear System Identification with Unknown Piecewise Time-Varying Delay

Identification of nonlinear systems with unknown piecewise time-varying delay is concerned in this paper.Multiple auto regressive exogenous(ARX) models are identified at different process operating points,and the complete dynamics of the nonlinear system is represented by using a combination of a normalized exponential function as the probability density function with each of the local models.The parameters of the local ARX models and the exponential functions as well as the unknown piecewise time-varying delays are estimated simultaneously under the framework of the expectation maximization(EM) algorithm.A simulation example is applied to demonstrating the proposed identification method.

Electro-Hydraulic Servo System Identification of Continuous Rotary Motor Based on the Integration Algorithm of Genetic Algorithm and Ant Colony Optimization

In order to increase the robust performance of electro-hydraulic servo system, the system transfer function was identified by the intergration algorithm of genetic algorithm and ant colony optimization(GA-ACO), which was based on standard genetic algorithm and combined with positive feedback mechanism of ant colony algorithm. This method can obtain the precise mathematic model of continuous rotary motor which determines the order of servo system. Firstly, by constructing an appropriate fitness function, the problem of system parameters identification is converted into the problem of system parameter optimization. Secondly, in the given upper and lower bounds a set of optimal parameters are selected to meet the best approximation of the actual system. And the result shows that the identification output can trace the sampling output of actual system, and the error is very small. In addition, another set of experimental data are used to test the identification result. The result shows that the identification parameters can approach the actual system. The experimental results verify the feasibility of this method. And it is fit for the parameter identification of general complex system using the integration algorithm of GA-ACO.

DPCM-based vibration sensor data compression and its effect on structural system identification

Due to the large scale and complexity of civil infrastructures, structural health monitoring typically requires a substantial number of sensors, which consequently generate huge volumes of sensor data. Innovative sensor data compression techniques are highly desired to facilitate efficient data storage and remote retrieval of sensor data. This paper presents a vibration sensor data compression algorithm based on the Differential Pulse Code Modulation (DPCM) method and the consideration of effects of signal distortion due to lossy data compression on structural system identification. The DPCM system concerned consists of two primary components: linear predictor and quantizer. For the DPCM system considered in this study, the Least Square method is used to derive the linear predictor coefficients and Jayant quantizer is used for scalar quantization. A 5-DOF model structure is used as the prototype structure in numerical study. Numerical simulation was carried out to study the performance of the proposed DPCM-based data compression algorithm as well as its effect on the accuracy of structural identification including modal parameters and second order structural parameters such as stiffness and damping coefficients. It is found that the DPCM-based sensor data compression method is capable of reducing the raw sensor data size to a significant extent while having a minor effect on the modal parameters as well as second order structural parameters identified from reconstructed sensor data.

System Identification of Controlled Pulse Keyhole Plasma Arc Welding Process

The development of closed-loop control systems is one of the most effective ways to improve the stability of the keyhole status during keyhole plasma arc welding （K-PAW）. Due to the disadvantages of the ＂one-pulse-one-keyhole＂ technology based on the conventional square current waveform, the controlled pulse welding current waveform is newly applied to control the keyhole open and close periodically. In order to realize the real-time control on the keyhole behavior with this advanced current waveform, welding experiments and system identification are conducted based on the classical control theory. One complete welding cycle can be divided into 3 periods. The keyhole establishing time is the most important time variable, which determines the keyhole behavior and welding process stability. At the same time, the averaged effiux plasma arc voltage during one pulse cycle can reflect the real keyhole dimension and status in a real-time manner. Therefore, two single-input-single-output （SISO） systems are proposed, in which keyhole establishing time and keyhole average dimension are taken as the system controlled variables respectively. Welding experiments are designed with the peak current varying randomly. Experiments show that the keyhole establishing time changes in an opposite direction to the varied peak current, and the averaged efflux plasma arc voltage varies with the same trend as the peak current. Based on the least squares technique and F test of classical system identification, second order difference equation for keyhole establishing time/peak current system and first order difference equation for keyhole average dimension/peak current system are obtained. It is proved that the calculated data by the two mathematical expressions are well matched with the measured data. The proposed research provides mathematical expressions and theoretical analysis to develop closed-loop systems for the controlled pulse K-PAW.

Outliers, inliers and the generalized least trinuned squares estimator in system identification

The least trimmed squares estimator (LTS) is a well known robust estimator in terms of protecting the estimate from the outliers. Its high computational complexity is however a problem in practice. We show that the LTS estimate can be obtained by a simple algorithm with the complexity 0( N In N) for large N, where N is the number of measurements. We also show that though the LTS is robust in terms of the outliers, it is sensitive to the inliers. The concept of the inliers is introduced. Moreover, the Generalized Least Trimmed Squares estimator (GLTS) together with its solution are presented that reduces the effect of both the outliers and the inliers. Keywords Least squares - Least trimmed squares - Outliers - System identification - Parameter estimation - Robust parameter estimation This work was supported in part by NSF ECS — 9710297 and ECS — 0098181.

A Novel Parsimonious Neurofuzzy Model Applied to Railway Carriage System Identification and Fault Diagnosis

In this paper, we suggest a novel parsimonious neurofuzzy model realized by RBFNs for railway carriage system identification and fault diagnosis. To overcome the curse of dimensionality resulting from high dimensional input variables, in our developed model the features extracted from the available observations are regarded as the input variables by adopting the higher-order statistics(HOS) technique. Such a constructed model is also applied to a practical railway carriage system, simulation results indicate that the developed neurofuzzy model possesses strong identification and fault diagnosis ability.

Close-Loop System Identification Using Over-sampling Scheme and Its Estimate Accuracy Analysis

A new identification method for a linear discrete-time closed-loop system is proposed based on an output over-sampling scheme. When the system outputs are over-sampled the new output sequences would contain more information about the plant structure. Using general least squares method （GLS） the plant over-sampled model should be recognized. Then the original plant model should be obtained by its relationship with the over-sampled model. Compared with conventional approaches the advantage of the new method is that even if the ordinary identifiability conditions are not satisfied, a close-loop system can be identified by using the oversampled output without utilizing any external test signal. Accuracy analysis shows the relationship between the estimation error and the over-sampling rate. Numerical simulation illnstrates its effectiveness.

Research of internet worm warning system based on system identification

The frequent explosion of Internet worms has been one of the most serious problems in cyberspace security. In this paper, by analyzing the worm＇s propagation model, we propose a new worm warning system based on the method of system identification, and use recursive least squares algorithm to estimate the worm＇s infection rate. The simulation result shows the method we adopted is an efficient way to conduct Internet worm warning.

DYNAMIC COMPENSATION OF MEASURING SYSTEMS BY SYSTEM IDENTIFICATION APPROACH

A digital filtering method is presented to compensate the dynamic characteristics of measuring systems.The compensation filter has an infinite impulse response property and is designed by system identification approach from the known input output pairs of the measuring system.Applications of this method to eliminating the distortions of measured waveform in transient pulse measurement are investigated.Experimental results show that the measurement errors caused by the sensor are reduced to be very small after the use of the compensation filter.

An Improved Differential Evolution Trained Neural Network Scheme for Nonlinear System Identification

This paper presents an improved nonlinear system identification scheme using di？erential evolution （DE）, neural network （NN） and Levenberg Marquardt algorithm （LM）. With a view to achieve better convergence of NN weights optimization during the training, the DE and LM are used in a combined framework to train the NN. We present the convergence analysis of the DE and demonstrate the efficacy of the proposed improved system identification algorithm by exploiting the combined DE and LM training of the NN and suitably implementing it together with other system identification methods, namely NN and DE＋NN on a number of examples including a practical case study. The identification results obtained through a series of simulation studies of these methods on different nonlinear systems demonstrate that the proposed DE and LM trained NN approach to nonlinear system identification can yield better identification results in terms of time of convergence and less identification error.

Improved System Identification Approach Using Wavelet Networks

A new approach is proposed to improve the general identification algor ithm of multidimensional systems using wavelet networks. The general algorithm i nvolves mapping vector input into its norm to avoid problem of dimensionality in construction multidimensional wavelet basis functions. Thus, the basis function s are spherically symmetric without direction selectivity. In order to restore t he direction selectivity, the improved approach weights the input variables befo r e mapping it into a scalar form. The weights can be obtained using universal opt imization algorithms. Generally, only local optimal weights are obtained. Even s o, performance of identification can be improved.