Zhao Xiangmo

A New Adaptive Multistage Median Filter

A new filter structure called adaptive multistage median filter (say AMM for simple) is proposed. It follows the basic architecture of the multistage median filter (MLM), but its subfilters are based on the date-dependent and sizes variant subwindows, that allows AMM give better performance than MLM. In addition, like MLM, the proposed filter is not only computationally effective, but also needs no a priori information about image or noise. Experiment results are given to show the superiority of the new proposed filters.

FPGA implementation of VSCS-LMS algorithm for MEMS Gyroscope

In order to improve the precision of attitude estimation system, paper presents a signal extraction method of MEMS Gyroscope called VSCS-LMS algorithm based on FPGA. The design is established with the S-shaped function of sigmoid to reduce the sensitivity to noise of LMS algorithm. And its FPGA-based system structure is synthesized and simulated on Virtex II FPGA by reusing the single-order filter unit to complement the multi-order accumulation. The signal extraction simulation of MEMS gyroscope shows that it is reliable, high precision and high degree of real-time.

Auto anti-lock braking system bench test results classification model based on neural network

Auto anti-lock braking system (ABS) bench test is a safe, high-efficiency and low-cost method for ABS performance detection. The key parameters such as slip ratio, adhesion coefficient utilization rate and deceleration can be obtained quickly. In this paper, a classification model based on neural network for ABS bench test results was established. And the detailed BP network structure design process was presented. BP neural network self-learning ability was used to analyze the bench test data. A large number of model mapping relationships were summarized and stored in the networks. MATLAB simulations show that the BP neural network model can classify the ABS bench test results correctly for a variety of experimental conditions.

Application of EMD in the Vehicle Braking Performance Inspection System

To achieve the goal of full procedural detection of the changing of vehicle braking force, and to improve the accuracy and repeatability of the inspection system, after analyzing several commonly used data smoothing methods, we propose applying empirical modal decomposition to smoothing of car braking force data. Simulation and experimental results show that the method can weaken the effect caused by disturbance, and improve curve smoothness. It also can preserve the changing patterns of the original curve, which guarantees system detection accuracy and repeatability. This proves the effectiveness and feasibility of the proposed method.