Papers by Author: Jie Zhang

Abstract: A new unpowered decanter system was designed to lift the decanting sink only by the dynamic provided by liquid level difference. The system arranged the float tank pools symmetrically in both ends of the decanting sink. The system was run by the difference of liquid level between float tank pool and reaction pool. When the level difference got small, the float tank would rise and hence lift up the decanting sink in the reaction pool to up above the liquid level; when the level difference got large, the float tank no longer received the buoyance so the decanting sink would float in the reaction pool and decant. The testing results of the experiment device showed that by means of difference of liquid level to lift the decanting sink was feasible. The new system is simple in structure, so that the construction and maintenance cost is not high. It also does not consume electricity. It provides a new decanter model.

Abstract: In this paper, the Brown motion PSO is proposed to deal with the slow convergence, low precision and local optimal problem of the Particle Swarm Optimization (PSO) algorithm in solving the complex functions. The wave operator is designed, which is similar to the differential mutation operator, to improve the particle velocity formula. The new algorithm is applied in the design of the adaptive filter; Experiments results show that the new algorithm has the faster convergence than the traditional PSO algorithm, and it has the better stability.

Abstract: Particle swarm optimization (PSO) is a kind of evolutionary computation technology which simulates the behavior of biological species. The essence of adaptive noise cancellation (ANC) is adjust the weight value of filter based on the input signals, the LMS algorithm is commonly used in this system, However, the convergence behavior and maladjustment of the LMS algorithm is seriously affected by the step-size μ, and the optimum value of μ cannot be determined easily, In this paper, Particle Swarm Optimization with linear decreasing inertia weight is proposed to solve the filter problem instead of LMS, taking the FIR filter of ANC as example, the simulation shows that ANC based on the PSO algorithm is better than classic ANC based on the LMS algorithm, and it gives the satisfactory results.