In this paper, a new Multi-Layer Perceptron Neural Network (MLP NN) classifier is proposed for classifying sonar targets and non-targets from the acoustic backscattered signals. Besides the capabilities of MLP NNs, it uses Back Propagation (BP) and Gradient Descent (GD) for training; therefore, MLP NNs face with not only impertinent classification accuracy but also getting stuck in local minima as well as lowconvergence speed. To lift defections, this study uses Adaptive Best Mass Gravitational Search Algorithm (ABGSA) to train MLP NN. This algorithm develops marginal disadvantage of the GSA using the bestcollected masses within iterations and expediting exploitation phase. To test the proposed classifier, this algorithm along with the GSA, GD, GA, PSO and compound method (PSOGSA) via three datasets in various dimensions will be assessed. Assessed metrics include convergence speed, fail probability in local minimum and classification accuracy. Finally, as a practical application assumed network classifies sonar dataset. This dataset consists of the backscattered echoes from six different objects: four targets and two non-targets. Results indicate that the new classifier proposes better output in terms of aforementioned criteria than whole proposed benchmarks.
The paper considers an algorithm for increasing the accuracy of measuring systems operating on moving objects. The algorithm is based on the Kalman filter. It aims to provide a high measurement accuracy for the whole range of change of the measured quantity and the interference effects, as well as to eliminate the influence of a number of interference sources, each of which is of secondary importance but their total impact can cause a considerable distortion of the measuring signal. The algorithm is intended for gyro-free measuring systems. It is based on a model of the moving object dynamics. The mathematical model is developed in such a way that it enables to automatically adjust the algorithm parameters depending on the current state of measurement conditions. This makes possible to develop low-cost measuring systems with a high dynamic accuracy. The presented experimental results prove effectiveness of the proposed algorithm in terms of the dynamic accuracy of measuring systems of that type.
In this paper, the adaptive control based on symbolic solution of Diophantine equation is used to suppress circular plate vibrations. It is assumed that the system to be regulated is unknown. The plate is excited by a uniform force over the bottom surface generated by a loudspeaker. The axially-symmetrical vibrations of the plate are measured by the application of the strain sensors located along the plate radius, and two centrally placed piezoceramic discs are used to cancel the plate vibrations. The adaptive control scheme presented in this work has the ability to calculate the error sensor signals, to compute the control effort and to apply it to the actuator within one sampling period. For precise identification of system model the regularized RLS algorithm has been applied. Self-tuning controller of RST type, derived for the assumed system model of the 4th order is used to suppress the plate vibration. Some numerical examples illustrating the improvement gained by incorporating adaptive control are demonstrated.
Active Noise Control (ANC) of noise transmitted through a vibrating plate causes many problems not observed in classical ANC using loudspeakers. They are mainly due to vibrations of a not ideally clamped plate and use of nonlinear actuators, like MFC patches. In case of noise transmission though a plate, nonlinerities exist in both primary and secondary paths. Existence of nonlinerities in the system may degrade performance of a linear feedforward control system usually used for ANC. The performance degradation is especially visible for simple deterministic noise, such as tonal noise, where very high reduction is expected. Linear feedforward systems in such cases are unable to cope with higher harmonics generated by the nonlinearities. Moreover, nonlinearities, if not properly tackled with, may cause divergence of an adaptive control system. In this paper a feedforward ANC system reducing sound transmitted through a vibrating plate is presented. The ANC system uses nonlinear control filters to suppress negative effects of nonlinearies in the system. Filtered-error LMS algorithm, found more suitable than usually used Filtered-reference LMS algorithm, is employed for updating parameters of the nonlinear filters. The control system is experimentally verified and obtained results are discussed.
This paper presents an adaptive particle swarm optimization (APSO) based LQR controller for optimal tuning of state feedback controller gains for a class of under actuated system (Inverted pendulum). Normally, the weights of LQR controller are chosen based on trial and error approach to obtain the optimum controller gains, but it is often cumbersome and tedious to tune the controller gains via trial and error method. To address this problem, an intelligent approach employing adaptive PSO (APSO) for optimum tuning of LQR is proposed. In this approach, an adaptive inertia weight factor (AIWF), which adjusts the inertia weight according to the success rate of the particles, is employed to not only speed up the search process but also to increase the accuracy of the algorithm towards obtaining the optimum controller gain. The performance of the proposed approach is tested on a bench mark inverted pendulum system, and the experimental results of APSO are compared with that of the conventional PSO and GA. Experimental results prove that the proposed algorithm remarkably improves the convergence speed and precision of PSO in obtaining the robust trajectory tracking of inverted pendulum.
There are many industrial environments which are exposed to a high-level noise, sometimes much higher than the level of speech. Verbal communication is then practically unfeasible. In order to increase the speech intelligibility, appropriate speech enhancement algorithms can be used. It is impossible to filter off the noise completely from the acquired signal by using a conventional filter, because of two reasons. First, the speech and the noise frequency contents are overlapping. Second, the noise properties are subject to change. The adaptive realisation of the Wiener-based approach can be, however, applied. Two structures are possible. One is the line enhancer, where the predictive realisation of the Wiener approach is used. The benefit of using this structure it that it does not require additional apparatus. The second structure takes advantage of the high level of noise. Under such condition, placing another microphone, even close to the primary one, can provide a reference signal well correlated with the noise disturbing the speech and lacking the information about the speech. Then, the classical Wiener filter can be used, to produce an estimate of the noise based on the reference signal. That noise estimate can be then subtracted from the disturbed speech. Both algorithms are verified, based on the data obtained from the real industrial environment. For laboratory experiments the G. R. A. S. artificial head and two microphones, one at back side of an earplug and another at the mouth are used.